Daemons & Ghosts

=><= #[[Daemons & Ghosts->Quote]] An Auto-Reflexive Digital Logbook on AI-Augmented Creative Practice <img src="artvoice.jpeg" class="ribbon"/> Links to internal passages look [[like this->The First Post]]. Links to external sites look <a href="http://www.thedigitalreview.com"; target="_blank">like this</a>. [[Start Reading Here->The First Post]]

=><= 07 November 2023: The First Entry <== I recently enrolled at <a href="https://www.aru.ac.uk"; target="_blank">ARU</a> as a postgraduate researcher, working towards a PhD in creative writing. I’m intending to produce a linked collection of stories, so when I came across a <a href="https://thedigitalreview.com/submissions.html"; target="_blank">submission call</a> for “born-digital works that delve into the concept of machine-augmented creativity” it seemed like a good time to create an AI-assisted story. In May I attended Lynda Clark’s online workshop <a href="https://www.nawe.co.uk/DB/events/ai-for-writers-with-dr-lynda-clark.html"; target="_blank">AI for Writers, or: Creativity Amplification</a>, which gave me an introduction to some of the AI tools available for writers. I may also use her <a href="https://github.com/IngameDundee/CreativityAmplification/wiki/Quickstart-Guide"; target="_blank">Toolkit</a> as a resource. I will use <a href="https://chat.openai.com"; target="_blank">ChatGPT</a> as a ‘prompter’ at the planning stage, <a href="https://textsynth.com/completion.html"; target="_blank">textsynth</a> as a ‘co-creator’ to aid the writing, and <a href="http://www.artbreeder.com"; target="_blank">artbreeder</a> to provide illustrations. The submission call specifies born-digital works, which include interactivity - so for this I’m going to produce an online interactive story in <a href="http://www.twinery.org"; target="_blank">Twine</a>, which I can publish as html. For this exercise I’m going to plan and write a story I’ve not developed yet at all: my ‘Easter Egg’ story. This is the one that’s going to sit outside of the branches of my linked story collection, for ‘cheating’ readers to find (interactive novels like <a href="https://www.goodreads.com/en/book/show/279985"; target="_blank">//Life's Lottery//</a> often include these). This means I'm free to play and experiment with this story, and see where the AI influence takes it - but I can use the parent and child characters from the other stories as a starting point. Using AI as a tool I’m going to create a ‘dream’ story. The ‘you’ character has been pumped full of hormones in order to prepare them for parental bonding with the new baby. This gives them strange dreams when they sleep, dreams that will foreshadow some elements of the collection. So I intend to create a dreamscape on Twine that the reader can navigate. However, this collaborative exercise makes me a little uncomfortable. As long as the initial ideas and characters are mine, I should be able to claim the AI-assisted story as mine too. Or should I? What is ‘mine’? Writers' ideas may come from memories, something they’ve experienced or read about, been told about or stumbled across... Consciously or not, we all draw ideas from somewhere. Is a writer no different to an AI chatbot, absorbing and regurgitating stories from life experience the way AI does from the internet? Unless I’m wrong, and that’s not where inspiration really comes from. Perhaps, as Elizabeth Gilbert discussed in her <a href="https://www.ted.com/talks/elizabeth_gilbert_your_elusive_creative_genius?language=en"; target="_blank">TED Talk</a> in 2009, creativity comes from an external source (or seems to), like the daemons of the ancient world. While she may have been talking about a ‘psychological construct’ to protect the writer from their work, the concept of daemon may be particularly relevant here. <a href="https://www.techtarget.com/whatis/definition/daemon"; target="_blank">In computing</a> a daemon is a computer process that runs in the background. Perhaps AI chatbots can also function as a sort of daemon, whispering ideas in a writer's ear, producing surprising and surreal fiction seemingly from nowhere. Perhaps I feel uncomfortable utilising AI tools for my writing because I don’t want to replace or bypass my own personal ‘daemon’, and use AI as an artificial one. Could that really produce a meaningful text? Questions I want to ask with this fictional experiment: Can an AI-assisted story be meaningful? Will it still be my story? How much control will I have? And does it matter if together we produce a good story? I am going to take Lynda Clark’s advice and treat the AI as a tool or co-author, and as suggested in R. Lyle Skains’ <a href="https://www.tandfonline.com/doi/full/10.1080/14682753.2017.1362175"; target="_blank">Creative Practice as Research: Discourse on Methodology</a> (2018), I'll keep a research log. I’ll record every question and response. [[First Experiments->4 Experiments]] Skip straight to the [[last post->Conclusion]].

=><= 07 November 2023: Experiment 1, using ChatGPT 3.5 <== > Hello. Can you help me to [[plan a story?]]

=><= 08 November 2023: Failing Drafts <== I was going to edit the AI-produced drafts today, but the writing feels clunky and empty, and I don’t feel inspired to turn it into anything better. I wonder if I should start again, try to rephrase my questions and supply different starting material. I have conducted some further experiments. When given a single line to begin a story, ChatGPT produced an odd but interesting [[piece of fiction->story by chat]]. Why is it that it seems to produce better work, when given less input from the human collaborator? I also experimented with asking for a [[Twine story]]. This felt almost like playing one, as ChatGPT would continue along whichever choice I made. [[Next Entry->control]]

{ (enchant: 'ghost', (text-colour: grey)+ (text-style: 'smear')) (enchant: 'daemon', (text-colour: red)+ (text-style: 'smear')) (enchant: 'Ghost', (text-colour: grey)+ (text-style: 'smear')) (enchant: 'Daemon', (text-colour: red)+ (text-style: 'smear')) (enchant: 'Ghosts', (text-colour: grey)+ (text-style: 'smear')) (enchant: 'Daemons', (text-colour: red)+ (text-style: 'smear')) (enchant: 'ghosts', (text-colour: grey)+ (text-style: 'smear')) (enchant: 'daemons', (text-colour: red)+ (text-style: 'smear')) (enchant: 'daemonology', (text-colour: red)+ (text-style: 'smear')) (enchant: 'golem', (text-colour: orange)+(text-style: 'smear')+(text-style: 'bold')) (enchant: 'golems', (text-colour: orange)+(text-style: 'smear')+(text-style: 'bold')) (enchant: 'Golem', (text-colour: orange)+(text-style: 'smear')+(text-style: 'bold')) (enchant: 'Golems', (text-colour: orange)+(text-style: 'smear')+(text-style: 'bold')) (enchant: ?Page, (font: "calibri")) }

Of course, I'd be happy to help you plan a story! Whether you need assistance with brainstorming ideas, developing characters, creating a plot outline, or any other aspect of storytelling, just let me know what you're looking for, and I'll do my best to assist you. Please provide me with some initial details or ideas you have for your story, and we can build on that foundation. > It’s a short story, with a new parent as the main character. They are exhausted and hormonal, and slip into a strange half-dream. In the dream there is a strange creature, a sort of daemon, lurking somewhere in the house but they can’t find it. They fear for their new baby. [[What should happen next?]] Skip to [[reflections->reflections1]].

That's a great starting point for a short story. Let's continue to develop it. Here's a possible outline for the next part of your story: Introduce the New Parent: Begin by delving deeper into the thoughts and feelings of the new parent. Describe their exhaustion and hormonal fluctuations, making the readers empathize with their state of mind. Transition to the Dream: As the parent falls asleep or slips into their half-dream, transition into the dream sequence. Describe the dream as a surreal and eerie experience. Use vivid imagery to convey the strangeness of the daemon and the house. Search for the Daemon: The parent's primary concern in the dream is the safety of their new baby. They become convinced that the daemon is a threat to the baby's well-being. Describe their frantic search for the creature, creating tension and a sense of urgency. Reveal Clues and Mystery: As the parent searches the dream house, drop subtle hints and clues that keep the reader guessing. Is the daemon real or a manifestation of their sleep-deprived mind? Use symbolism and ambiguity to keep the mystery alive. Confrontation or Revelation: The dream should build to a climax. The parent either confronts the daemon or uncovers some kind of revelation. This can be a moment of personal growth or realization. Return to Reality: Gradually bring the parent back to the waking world. They might wake up in a cold sweat, disoriented and even more exhausted than before. The transition between the dream and reality should be seamless and jarring, blurring the lines between the two. Emotional Resolution: Conclude the story by addressing how the parent feels after this dream experience. Have they gained a new perspective on their role as a parent? Is there a sense of relief or lingering unease? End with a Twist or Ambiguity: You can choose to end the story with a twist or leave it somewhat open-ended, letting readers draw their conclusions about the nature of the dream and the daemon. Remember to weave in sensory details, emotions, and symbolism to create a rich and engaging narrative. This outline should give you a solid framework to build your short story around. Feel free to modify it to suit your creative vision. > That’s great, thank you. Could you please summarise a very simple plot line for me? [[Just the beginning, middle and end?->simple plot line]] Skip to [[reflections->reflections1]].

Certainly! Here's a simplified plot outline for your short story: Beginning: Introduce a new parent who is exhausted and hormonal, struggling with the demands of parenthood. The parent slips into a strange half-dream state. Middle: In the dream, the parent encounters a mysterious daemon lurking in their house. They search for the daemon, fearing for the safety of their new baby. The dream is filled with tension, clues, and ambiguity, keeping the parent and readers in suspense. End: The dream reaches a climax where the parent confronts the daemon or makes a significant revelation. The parent wakes up in the real world, disoriented and emotionally impacted. The story concludes with the parent reflecting on their experience and its impact on their feelings and perspective as a new parent. > Could we change this up a bit, to make it more unpredictable and surreal? Perhaps, the addition of some new characters? [[How about an electric sheep?->add electric sheep]] Skip to [[reflections->reflections1]].

Certainly, let's add some surreal and unpredictable elements to the story, including the introduction of new characters and an electric sheep. Here's an updated plot outline: Beginning: Introduce a new parent who is exhausted and hormonal, struggling with the demands of parenthood. The parent slips into a strange half-dream state, characterized by surreal and shifting realities. Middle: In the dream, the parent encounters a strange, otherworldly character, perhaps a quirky guide who claims to know the way to find the mysterious daemon. The parent and their guide embark on a surreal journey through a dreamlike house filled with bizarre, ever-changing rooms and landscapes. They are accompanied by an electric sheep, which emits sparks of light and seems to have a peculiar connection to the daemon. The dream world is filled with surreal and unpredictable challenges, keeping the parent and readers guessing about the true nature of the daemon. End: The dream reaches a climactic moment where the parent and their guide confront the daemon, and the electric sheep plays a pivotal role in the resolution. The parent wakes up in the real world, disoriented, and emotionally impacted, with the lingering question of whether their dream journey had any real-world significance. The story concludes with the parent reflecting on their dream experience and how it has affected their perception of parenthood and the challenges they face. This updated plot introduces more surreal and unpredictable elements while incorporating the electric sheep and new characters to enhance the dreamlike quality of the story. > [[Thank you.]]

You're welcome! If you have any more questions or need further assistance as you develop your story, feel free to ask. Good luck with your writing, and I hope your story turns out to be a captivating and unique piece of fiction! [[Reflections->reflections1]] [[Experiment 2-> GPT Exp2]]

I was a little underwhelmed by this process. The plots given used the elements I’d asked for, but in a very predictable and cliché way. However, some of the suggestions could have been useful if I had been confused as to how to proceed - using a guide character to lead the ‘parent’ to the daemon for example. With hindsight, the AI has not really added anything new or unexpected to the story, which means that even if I asked the AI to write the story for me, it might still feel like mine. However, I don’t intend to take that course, since I want this to be a collaboration. I want an AI to play the role of the daemon, almost. I’m going to try again, and ask different questions. I’m going to leave the initial summary of the story more open-ended, and see what happens. [[Go to the next experiment-> GPT Exp2]], or skip straight to [[textsynth experiment]].

=><= 07 November 2023: Experiment 2 <== > Hello. Can you help me to replan a story? Of course, I'd be happy to help you replan a story. Please provide me with some details about your story and what aspects you'd like to work on, such as the plot, characters, setting, or any specific issues you're facing, and I'll do my best to assist you. > It’s a short story, with a new parent as the main character. They are exhausted and hormonal, and slip into a strange half-dream. In the dream there is a strange creature, a sort of daemon, whispering in their ear. That sounds like an interesting concept for a short story. Let's work on replanning it. Here's a basic outline that you can use as a starting point: Title: (You can decide on a title that reflects the theme of the story.) Character: New Parent (Main Character) - struggling with exhaustion and hormonal changes. Setting: A cozy, dimly lit bedroom where the new parent is trying to get some rest. Plot: Introduction Introduce the new parent and their current state of exhaustion and hormonal imbalance. Set the scene in the bedroom. The Dream Begins The new parent starts to drift into a half-dream state. The Strange Creature Describe the appearance and behavior of the strange creature (daemon) that appears in the dream. The daemon begins whispering enigmatic, cryptic messages to the new parent. These messages may be related to their current struggles, fears, or anxieties. Conflict and Confusion The new parent is initially confused and alarmed by the presence of the daemon. The daemon's whispers become more intense, and the parent struggles to understand their significance. Exploration of Emotions Use the interactions with the daemon to explore the new parent's emotional state. Show how the parent's exhaustion and hormonal changes are affecting their mental and emotional well-being. Resolution The dream starts to fade as the parent awakens, but the messages from the daemon stay with them. The parent reflects on the dream and the messages, gaining some insight or clarity about their feelings and challenges. Conclusion End the story with the parent in the bedroom, perhaps with a sense of acceptance or understanding about their situation. Themes: Parenting, exhaustion, hormonal changes, dreams, self-discovery, the subconscious. You can expand on each of these plot points and add more details to create a compelling and evocative short story. Feel free to adjust the plot, characters, and setting to better fit your vision for the story. If you have any specific questions or need further guidance, please let me know. > That’s interesting, but I think I’d like more action in the story. Could the parent travel through the strange world of their dream? [[What might happen next?->more action]] Skip to [[reflections->reflections2]].

Certainly, we can add more action and adventure to the story by having the new parent travel through the strange world of their dream. Here's an updated outline: Title: (You can still decide on a title that reflects the revised theme.) Character: New Parent (Main Character) - struggling with exhaustion and hormonal changes. Setting: The dream world, a surreal and ever-changing landscape. Plot: Introduction Introduce the new parent and their current state of exhaustion and hormonal imbalance. Set the scene in the bedroom as the parent falls asleep. The Dream Journey Begins The new parent starts to drift into a half-dream state. In the dream, they find themselves in a strange and surreal landscape with bizarre creatures and unusual landscapes. Encounters in the Dream World The parent encounters various obstacles and challenges in the dream world. Each encounter in the dream world reflects a different aspect of their struggles as a new parent. They may encounter obstacles related to exhaustion, self-doubt, and hormonal changes. The Strange Daemon The daemon reappears in the dream world, continuing to whisper cryptic messages. The daemon serves as a guide or a mysterious figure who helps the parent navigate the dream world. Action and Conflict The parent and the daemon face challenges and adversaries in the dream world. These challenges may represent the fears and anxieties the parent has been dealing with in real life. Personal Growth and Revelation As the parent overcomes obstacles and interacts with the daemon, they begin to understand the meaning behind the daemon's messages. They experience personal growth and self-discovery. Awakening The parent awakens from the dream, but they now have a new perspective on their struggles and feelings as a new parent. Conclusion End the story with the parent in the bedroom, reflecting on their dream journey and the insights they've gained. They may feel more empowered and better equipped to face the challenges of new parenthood. Themes: Parenting, exhaustion, hormonal changes, dreams, self-discovery, personal growth, surreal adventure. This revised outline adds more action and adventure to the story, as the new parent traverses a surreal dream world, facing challenges and gaining insights. You can further develop the dream world, the specific challenges the parent encounters, and the nature of their interactions with the daemon to make the story engaging and dynamic. > I think the story needs more characters. Could their child be a toddler, lost in the dreamscape? Maybe they’re following a different guide, an [[electric sheep?->sheep again]] Skip to [[reflections->reflections2]].

Expanding the story with additional characters and a toddler in the dream world adds more complexity and intrigue. Here's an updated outline that includes these elements: Title: (You can still decide on a title that reflects the revised theme.) Characters: New Parent (Main Character) - struggling with exhaustion and hormonal changes. Toddler (Child) - lost in the dream world. Strange Daemon - the mysterious guide. Electric Sheep - another guide in the dream world, leading the toddler. Setting: The dream world, a surreal and ever-changing landscape. Plot: Introduction Introduce the new parent and their current state of exhaustion and hormonal imbalance. Set the scene in the bedroom as the parent falls asleep. The Dream Journey Begins The new parent starts to drift into a half-dream state. In the dream, they find themselves in a strange and surreal landscape with bizarre creatures and unusual landscapes. Discover that their toddler is also present, wandering in the dream world. Encounters in the Dream World The parent and toddler encounter various obstacles and challenges in the dream world. Each encounter in the dream world reflects different aspects of their struggles as new parents. They may encounter obstacles related to exhaustion, self-doubt, and hormonal changes. The Strange Daemon and Electric Sheep The daemon and electric sheep serve as guides to the parent and the toddler, respectively. Each guide imparts wisdom and guidance in their own unique way. Action and Conflict The parent, toddler, daemon, and electric sheep face challenges and adversaries in the dream world. These challenges may represent the fears and anxieties the parent has been dealing with in real life. Personal Growth and Revelation As the characters navigate the dream world, they begin to understand the meaning behind the guides' messages. They experience personal growth, self-discovery, and a deeper connection with their child. Awakening The parent awakens from the dream, but they now have a new perspective on their struggles and feelings as a new parent. They find their toddler peacefully asleep in the real world. Conclusion End the story with the parent feeling more empowered and better equipped to face the challenges of new parenthood. They have a stronger bond with their child, forged through their shared dream experience. Themes: Parenting, exhaustion, hormonal changes, dreams, self-discovery, personal growth, surreal adventure, family bonds. This expanded outline incorporates the toddler and the electric sheep as additional characters, making the story more dynamic and complex. The interactions and challenges in the dream world can symbolize the parent's journey through the early stages of parenthood and their evolving relationship with their child. > Have any stories like this been [[written already?]] Skip to [[reflections->reflections2]].

While I can't provide you with real-time information or access to specific stories written after my knowledge cutoff date in January 2022, I can tell you that stories with elements of surreal dreamscapes, parental struggles, and unique guides or characters are not uncommon in literature and creative writing. Many authors have explored similar themes, but the specific details, characters, and settings would vary from one story to another. Authors like Haruki Murakami often incorporate dreamlike elements and surreal landscapes into their works, creating unique and thought-provoking stories. However, it's the execution, style, and the unique voice of the writer that make each story distinctive. If you're concerned about the originality of your story, it's always a good idea to write it in your own voice and style, infusing it with your personal experiences and insights. Your unique perspective and creativity will make your story stand out, even if it shares some thematic elements with existing works. You can also consider doing a search or consulting a literary database to find more recent stories with similar themes if you're looking for direct comparisons or sources of inspiration. > Are there any childrens' stories or folktales that have a [[similar plot?]] Skip to [[Reflections->reflections2]]

There are several children's stories and folktales that feature elements of dreamlike journeys and encounters with strange characters. While these may not align perfectly with the plot you've outlined, they share some thematic similarities. Here are a few examples: Alice's Adventures in Wonderland by Lewis Carroll: This classic children's story follows Alice as she falls into a fantastical and dreamlike world where she encounters peculiar characters and experiences strange events. Where the Wild Things Are by Maurice Sendak: In this beloved children's book, Max embarks on a journey to a mysterious island where he meets wild creatures. It explores themes of imagination and self-discovery. Peter Pan by J.M. Barrie: Peter Pan takes Wendy, John, Michael, and the Lost Boys on magical adventures to Neverland, a world of imagination and fantastical creatures. The Wizard of Oz by L. Frank Baum: Dorothy's journey to the magical land of Oz involves strange and whimsical characters as she seeks to find her way back home. The Little Prince by Antoine de Saint-Exupéry: While not exactly a folktale, it's a story with dreamlike elements and philosophical themes. The Little Prince travels to different planets, each inhabited by unique characters. While these stories may not directly match your plot, they do share elements of fantastical journeys, encounters with unusual characters, and exploration of inner worlds or emotions, which align with your story's themes. You can draw inspiration from these classics and adapt them to suit your own narrative. > What about fairytales? Fairy tales are rich sources of inspiration and often contain elements of dreams, strange creatures, and inner journeys. While there may not be fairy tales with plots identical to the one you've outlined, here are a few fairy tales that contain elements that could serve as inspiration for your story: "The Snow Queen" by Hans Christian Andersen: In this fairy tale, the protagonist Gerda embarks on a journey to rescue her friend Kay, who has been captured by the Snow Queen. Along her journey, she encounters various strange characters and challenges. "The Golden Bird" (Grimm Brothers): This tale features a magical bird, a golden apple tree, and a series of adventures that the protagonist goes through in a dreamlike quest. "The Red Shoes" (Hans Christian Andersen): This tale tells the story of Karen, who becomes entranced by a pair of red shoes. The shoes take on a life of their own, and she goes through a series of bizarre and dreamlike experiences. "Thumbelina" (Hans Christian Andersen): Thumbelina goes on a journey through various miniature worlds, meeting unique characters and creatures along the way. While none of these fairy tales align precisely with your plot, they contain elements of surreal adventures, encounters with strange characters, and inner journeys that could provide inspiration for your story. You can draw from these tales to create a unique and original story with the elements you have in mind. Fairy tales often serve as excellent sources of inspiration for new and inventive narratives. > How could I make my story more like a fairytale or folktale? To make your story more like a fairytale or folktale, you can infuse it with the characteristic elements and style often found in these genres. Here are some suggestions to give your story that fairytale or folktale quality: Timeless and Universal Themes: Fairytales and folktales often explore timeless themes such as good vs. evil, the hero's journey, and the consequences of one's actions. Make sure your story has a strong underlying theme that resonates with readers of all ages. Simplistic Language and Style: Use a simple and straightforward narrative style that's easy to follow. Avoid overly complex language, and focus on storytelling that captures the imagination. Moral or Lesson: Incorporate a moral or lesson into your story. Many fairytales and folktales convey a moral message, and characters' actions often lead to consequences that reflect these morals. Enchanting Setting: Create an enchanting and vivid setting for your dream world. Describe the dream world with magical and fantastical imagery that evokes a sense of wonder. Archetypal Characters: Develop characters that are archetypal and easily recognizable, such as the hero, the mentor, the trickster, or the villain. These archetypes often appear in traditional tales and resonate with readers. Magical Elements: Introduce magical or supernatural elements into the story. These could be mythical creatures, enchanted objects, or spells that add to the sense of wonder and magic. Repetition and Patterns: Incorporate repetition or patterns in your story. Traditional tales often use repetition to create a rhythmic and memorable narrative. Traditional Story Structure: Consider using a traditional story structure with clear exposition, rising action, climax, and resolution. This structure helps create a sense of familiarity. Illustrations: If possible, include illustrations or imagery that complements the storytelling and adds to the fairytale quality of your story, especially if you're creating a children's story. Narration or Framing: You can use a storytelling device where the main story is framed within a narrative, like "Once upon a time" or "In a land far, far away." This helps create a timeless and fantastical atmosphere. Simplicity in Morality: Keep the moral of your story simple and clear. Fairytales and folktales often teach straightforward life lessons. Happy Endings: Traditional fairytales often have happy endings, where good triumphs over evil and characters are rewarded. Consider giving your story a satisfying and uplifting conclusion. Remember that while you can incorporate these elements, it's essential to maintain your unique voice and perspective. Your story can be a blend of traditional fairytale elements and modern storytelling techniques, creating something both familiar and fresh for your readers. > How could I reframe the story as a fairytale in the style of the Brothers Grimm? To reframe your story as a fairytale in the style of the Brothers Grimm, you'll want to adopt their distinctive narrative elements, tone, and structure. The Brothers Grimm are known for their dark and sometimes gruesome tales, often filled with moral lessons. Here's how you can adapt your story: Dark and Moralistic Themes: Infuse your story with dark and moralistic themes. Brothers Grimm tales frequently involve hardship, consequences for one's actions, and sometimes grim outcomes. Grim Setting: Create a setting that is somewhat eerie or unsettling. The forest is a common backdrop in Grimm tales and can add a sense of foreboding. Simplistic Language: Use simple and straightforward language. Brothers Grimm stories are known for their plain, direct narrative style. Mysterious and Macabre Characters: Introduce mysterious and sometimes macabre characters. These could be the daemon, otherworldly creatures, or even humans with dark intentions. Moral Lessons: Ensure that your story conveys a clear moral lesson. Grimm tales often feature characters who make choices and face consequences, so the moral of the story is evident. Punishments and Rewards: Include punishments for wrongdoing and rewards for virtuous behavior in your story. These reinforce the moral aspects of the tale. Supernatural Elements: Integrate supernatural elements like magic, curses, and enchanted objects. These add to the fairytale mystique. Rhythmic Repetition: Incorporate repetition and patterns in your storytelling. Grimm tales often use repetition to create a rhythmic and memorable narrative. Clear Story Structure: Follow a clear and structured story pattern. Traditional fairytales typically have a recognizable structure with exposition, rising action, climax, and resolution. Folklore Elements: Draw inspiration from folklore and legends. The Brothers Grimm often adapted and collected stories from German folklore. Tragic and Unpredictable Outcomes: Consider a tragic or unexpected twist in your story. Grimm tales are known for their endings that are not always conventionally happy. Narrative Framing: Begin your story with a traditional narrative frame such as "Once upon a time" or "In a land far, far away." By incorporating these elements, you can give your story the distinctive style and tone of a Brothers Grimm fairytale. Keep in mind that while you draw from this style, your unique storytelling voice and your own creative elements should shine through to create a memorable and original tale. > Can you give me a basic plotline, just a beginning, middle and end, for the story as a gothic science fiction tale? Certainly! Here's a basic plotline for your story as a gothic science fiction tale: Title: (You can decide on a title that reflects the gothic science fiction theme.) Beginning: Introduce the new parent who is struggling with exhaustion and hormonal changes. They have recently moved to a remote, isolated mansion with their toddler to escape the chaos of the city. The parent's sleep-deprived state makes them susceptible to vivid and unsettling dreams. They start experiencing recurring dreams in which they are beckoned by a strange and ethereal presence. This presence, which appears as a spectral figure, urges the parent to explore the mansion's hidden laboratory, which is filled with mysterious, old scientific equipment and an ominous, antique computer system. Middle: As the parent delves deeper into the laboratory, they discover that the mansion's previous occupant was a scientist obsessed with conducting bizarre experiments, seeking to unlock the secrets of the human mind and consciousness. The parent stumbles upon a series of old journal entries and experiment logs that reveal the former occupant's descent into madness and their attempts to bridge the gap between dreams and reality. Meanwhile, the toddler, still lost in the dreamscape, begins to exhibit strange behaviors, seemingly influenced by the spectral figure from the parent's dreams. End: The parent's exploration of the laboratory leads them to a shocking revelation: the spectral figure they've been encountering in their dreams is a manifestation of the consciousness of the mansion's former owner, trapped within the mansion's computer system. The parent must confront the vengeful spirit of the scientist and find a way to release the toddler from the dream world. In a climactic confrontation within the laboratory, the parent makes a heart-wrenching decision that ultimately allows the toddler to return to the real world, while the vengeful spirit finds a form of release or closure. The story ends with a sense of unease and uncertainty as the parent and child, forever changed by their experiences, leave the mansion, leaving behind the haunted laboratory and its mysteries. This plotline combines elements of gothic horror and science fiction, with a focus on the eerie and the unexplained, and delves into the consequences of scientific experimentation on the human psyche. You can further develop the plot and add specific details to make it a compelling and chilling gothic science fiction tale. > Brilliant, thank you! NetworkError when attempting to fetch resource. [[Reflections->reflections2]]

I don't know what caused ChatGPT's error... but otherwise this was much better. Once I asked for a plot in the style of a gothic science fiction, the AI added lots of new details to the story that I hadn’t specifically asked for. Presumably, this is because it could draw on its information on gothic tales. For my third attempt, I’m going to give it my basic story elements and ask for a plot in the style of Black Mirror. By limiting it more than simply asking for ‘gothic’, I am concerned that it might give me elements from actual episodes - but I won’t know unless I ask. As an aside, being able to ask about similar stories produced some very interesting and - hopefully - useful answers. This AI might be more useful as a research tool than a provider of creative inspiration… [[Go to the next experiment-> GPT Exp3]], or skip straight to [[textsynth experiment]].

=><= 07 November 2023: Experiment 3 <== > Hello. Can you help me to outline a story? It’s a short story, with a new parent as the main character. They are exhausted and hormonal, and slip into a strange half-dream. In the dreamscape, they have to protect their toddler from strange characters, like an impish daemon and an electric sheep. I’d like to produce a plot in the style of a Black Mirror episode. Title: "Dreamscape Guardian" Setting: The story is set in a near-future world where advanced technology is integrated into daily life, resembling the eerie and futuristic ambiance of a "Black Mirror" episode. Main Character: Name: Sarah Profile: A new parent, struggling with exhaustion and postpartum hormones. Plot Outline: Act 1: The Exhausted New Parent Introduce Sarah, a new mother, struggling with the overwhelming demands of parenthood. She's sleep-deprived and feeling isolated. Act 2: The Dreamworld Sarah finally falls into a deep sleep after days of exhaustion and hormonal fluctuations. In her dream, she finds herself in a surreal and unsettling dreamscape, reminiscent of a "Black Mirror" episode. The world is distorted, and technology is omnipresent. Her toddler, Lily, is in this dream with her. She must protect Lily from the bizarre characters that inhabit the dreamscape. Introduce the impish daemon, a mischievous and malicious figure who torments Sarah and Lily. Another threat is the electric sheep, a robotic entity that's both mysterious and menacing, appearing almost everywhere, emitting unsettling electrical sounds. Act 3: Protecting Lily Sarah navigates the dreamworld, using the rules of the dreamscape to protect her child. She's challenged by bizarre obstacles and hostile characters. She discovers that in this dreamscape, her thoughts and emotions are connected to the environment, influencing the reality of the dream. As she struggles to protect Lily, she learns to harness her maternal instincts and emotional strength to overcome the obstacles. The electric sheep and impish daemon become more menacing and persistent, representing the anxiety and challenges of parenthood. Act 4: Resolution Sarah's determination to protect Lily helps her to gradually gain control of the dreamscape, bending it to her will. She confronts the impish daemon and the electric sheep, representing her inner demons and fears as a new parent. The dreamscape starts to unravel, and Sarah wakes up in the real world, feeling more empowered and capable. She now has a newfound confidence in her abilities as a parent, having conquered the challenges in her dream. Conclusion: "Dreamscape Guardian" is a psychological journey that explores the emotional struggles of new parenthood and the power of a mother's love and determination. The story incorporates the eerie, futuristic elements of a "Black Mirror" episode to create a compelling and thought-provoking narrative. > Are there any episodes [[like this already?]] Skip to [[reflections->reflections3]].

While the concept of a parent navigating a surreal dreamscape to protect their child is unique, there are some "Black Mirror" episodes and other media that explore elements of technology, dreams, and psychological challenges. Here are a few examples that touch on similar themes: "Black Mirror" Episodes: "White Christmas" (Season 2, Episode 4) features a story about digital consciousness, isolation, and the consequences of technology. While it doesn't involve a parent-child dynamic, it explores psychological challenges and the eerie atmosphere typical of the series. "Inception" (2010): This film, directed by Christopher Nolan, delves into dreamscapes, manipulation of reality, and the emotional dynamics between characters. While it doesn't directly involve new parenthood, it explores the power of the subconscious mind. "The Babadook" (2014): This psychological horror film touches on the themes of exhaustion and the emotional turmoil of parenting. The story involves a mother and her young son who face a sinister presence that emerges from a mysterious book. "The Twilight Zone" (Original Series and Reboots): Various episodes of "The Twilight Zone" explore dreamlike scenarios, technological dilemmas, and psychological challenges. It's a classic source for thought-provoking narratives. While there might not be an exact match to your story idea, these examples can serve as sources of inspiration and provide insights into how similar themes have been explored in different contexts. Your story can still be unique by combining these elements in a new and creative way. > Is there any published fiction that is similar? While your specific story concept involving a new parent navigating a surreal dreamscape to protect their child is relatively unique, there are works of published fiction that explore elements of dreams, surrealism, and psychological challenges. Here are a few examples that share thematic elements with your concept: "The Ocean at the End of the Lane" by Neil Gaiman: This novel blends reality with a dreamlike, otherworldly atmosphere. It follows the story of a man who reflects on his childhood and a supernatural encounter that occurs at a farm he visits. "The Shining" by Stephen King: This classic horror novel explores the psychological challenges of isolation, supernatural occurrences, and the impact on a family. While not centered on a dreamscape, it delves into the surreal and the psychological. "Alice's Adventures in Wonderland" by Lewis Carroll: This timeless classic follows Alice's adventures in a fantastical, dreamlike world. While not focused on parenthood, it does feature a surreal dreamscape and bizarre characters. "A Monster Calls" by Patrick Ness: This novel deals with themes of grief, loss, and coping with difficult emotions. The story revolves around a boy who encounters a monster in his dreams as he grapples with his mother's illness. "The Giver" by Lois Lowry: This dystopian novel explores the boundaries of reality, memory, and emotions. It follows a young boy in a seemingly utopian society who discovers the darker truths beneath the surface. These books incorporate elements of dreams, surrealism, and psychological challenges, and while they may not perfectly match your story concept, they can provide valuable inspiration and insights for your own creative work. > Do you know of any works written in the style of Black Mirror? "Black Mirror" is known for its unique style and themes that often explore the dark side of technology, society, and human nature. While there are no works exactly in the style of "Black Mirror," there are a few books and anthologies that capture the spirit of thought-provoking and dystopian science fiction similar to the series: "The Complete Stories of Philip K. Dick": Philip K. Dick's works often explore the intersection of technology, identity, and reality, creating a surreal and thought-provoking atmosphere that's akin to "Black Mirror." "The Twilight Zone" Companion by Marc Scott Zicree: While not fiction itself, this book delves into the classic TV series "The Twilight Zone," which has influenced "Black Mirror." It provides insights into the creation of episodes with similar themes and style. "Brave New World" by Aldous Huxley: This classic novel explores a future society driven by technology and conformity, with a focus on the dark side of an overreliance on technology and social control. "The Machine Stops" by E.M. Forster: This short story, written in 1909, envisions a future where humanity lives underground and depends on a vast, technology-driven Machine. It anticipates many themes seen in "Black Mirror." "The Test" by Sylvain Neuvel: This novella tells the story of an immigration test that takes a surreal and dark turn, examining themes of identity and human rights in a dystopian world. "The Water Will Come" by Jeff Goodell: While not fiction, this non-fiction book explores the very real consequences of climate change and the potential dystopian future it might bring, echoing some "Black Mirror" themes. These works capture the essence of speculative, thought-provoking, and often dystopian science fiction that "Black Mirror" is known for. While not identical in style or content, they explore similar themes and provoke reflection on the impact of technology and society on the human experience. > Thanks. You're welcome! If you have any more questions or need further assistance, feel free to ask. Happy writing, and best of luck with your story! [[Reflections->reflections3]]

The AI this time had enough information to give me a complete plot and even title the story, and fill in details like character names. There were more interesting ideas here, such as ‘She discovers that in this dreamscape, her thoughts and emotions are connected to the environment, influencing the reality of the dream’ but it’s not an original concept. Again, I think the most useful thing, that it seems to do best (the way I’m working with it anyway), is providing comparative titles. I am intrigued now to see what would happen if I asked it to write the story it plots - but I’m afraid to try that before I’ve written a version myself. I’m worried that facing a story already written would be a stumbling block for my own creative process. Comparing results afterwards however, might be more interesting. I’m going to change tack now… [[Go to the next experiment-> GPT Exp4]], or skip straight to [[textsynth experiment]].

=><= 07 November 2023: Experiment 4 <== > Hello. Can you help me to plan a story? I want to write a speculative short story about AI. A new parent is the main character. They are exhausted and hormonal, and slip into a strange half-dream. In the dreamscape, they have to protect their toddler from a daemon and an electric sheep. When they wake up they find themselves in a laboratory with a neural net attached to their skull. They learn that they were part of an experiment, and one of the figures they met was an AI designed to encourage the maternal bond. No one can account for the other. [[Response->reflections4]] [[Reflections->reflections4]]

There was no response as I never sent that last prompt, but aborted the experiment at this point - I realised I had enough to make a start with the next stage, writing the story with textsynth. Go to [[textsynth experiment]].

> Continue the story: The daemon tells you it can bring you ideas for fiction, that will make you the greatest writer the world has ever known. Intrigued by the daemon's proposition, I contemplated the possibilities it presented. The idea of becoming the greatest writer in the world had always been a dream of mine, but the road to achieving that status seemed insurmountable. Could this enigmatic daemon truly make it a reality? "Tell me more," I urged the daemon, my curiosity piqued. "How can you help me become the greatest writer in the world?" The daemon leaned in closer, its eyes gleaming with a hint of mischief. "I have the power to unlock your creativity, to channel the depths of your imagination, and bring forth ideas that will leave the world in awe. All I require in return is a simple pact. You must agree to write about the ideas I provide, no matter the consequences. Will you accept this deal?" The daemon's offer was tantalizing, but it came with an ominous undertone. I hesitated, realizing that such a pact might come with unforeseen and potentially dire consequences. As I weighed the pros and cons, I couldn't help but wonder how many other writers had faced this same decision throughout history. Sensing my hesitation, the daemon continued, "Consider this: How many writers would kill for the opportunity to craft stories that transcend time and culture, stories that touch the hearts and minds of generations to come? With my guidance, your words will not only capture the essence of humanity but also reshape it." My thoughts raced. What was the cost of greatness? Would I be willing to pay the price, even if it meant treading down a dark and treacherous path? The daemon's proposal seemed to dangle the world's literary treasures before me, and the allure was almost irresistible. With a deep breath, I made my decision. "I accept your offer," I said, my voice quivering with a mixture of excitement and trepidation. A sly grin spread across the daemon's face, revealing a row of sharp teeth. "Very well," it hissed, its tone laced with an unsettling satisfaction. "The pact is sealed." As the words left my lips, I could feel a profound shift in my consciousness. It was as if a hidden wellspring of creativity had been unlocked within me. Ideas flooded my mind, each more brilliant and captivating than the last. Characters, plots, and worlds unfolded before my inner eye, and I couldn't wait to put them down on paper. Over the weeks and months that followed, I wrote furiously, producing works that garnered praise and adoration from critics and readers alike. My stories touched the hearts of millions and seemed to resonate with the human condition in ways I could never have imagined. I had, indeed, become the greatest writer the world had ever known. Yet, as the years passed, I couldn't shake the nagging feeling that I had made a deal with a force beyond my understanding. The daemon's influence began to manifest in subtle and unsettling ways. My stories grew darker, more twisted, and more controversial. I found myself unable to resist the daemon's will, even when it led me down a morally dubious path. I was torn between the success I had achieved and the price I was paying for it. I had to confront the true nature of the pact I had made, and the inevitable reckoning that lay ahead. >Please Continue As the years went by, the weight of the daemon's influence became increasingly burdensome. The stories I wrote, while garnering acclaim and success, began to take a toll on my own well-being. I found myself consumed by the darkness that pervaded the narratives, and the once-boundless joy of creation turned into a haunting obsession. The world hailed me as a literary genius, yet I felt like a puppet, dancing to the daemon's malevolent tune. The characters in my stories took on a life of their own, their motivations and actions shaped more by the daemon's desires than my own creative instincts. It was as if I had become a mere vessel for the daemon's twisted visions. Haunted by the consequences of my pact, I sought refuge in solitude, hoping to break free from the insidious influence that clung to my every word. In the quiet of my study, I grappled with the moral dilemma that now defined my existence. Was the price of greatness worth the sacrifice of my own integrity? One fateful night, as I struggled with my inner demons, the daemon materialized before me, its presence casting a shadow over the room. "You cannot escape our pact," it whispered, its voice a sinister echo. "Your destiny is intertwined with mine, and the stories I provide will continue to flow through you." Desperation and defiance welled up within me. "I never signed up for this!" I retorted, my voice trembling with a mix of anger and fear. "I wanted to be a great writer, not a pawn in your dark game." The daemon's laughter echoed through the room, chilling me to the bone. "Greatness comes at a cost, mortal. You may have achieved unparalleled success, but you are now beholden to the very forces that elevated you." Determined to break free, I delved into research, seeking ancient texts and forbidden knowledge that might hold the key to severing the pact. The journey took me to remote corners of the world, where I encountered mystics and scholars with cryptic insights into the supernatural. As I unearthed arcane secrets, I realized that the only way to break the daemon's hold was to confront it directly. Armed with newfound knowledge and a resolve born of desperation, I performed a ritual in the depths of a forgotten temple, calling upon forces that rivaled the daemon's dark power. The air crackled with energy as the daemon materialized once more, its form contorting in anger. "You dare challenge me?" it bellowed, its voice a tempest of fury. With conviction in my heart, I spoke words of liberation, demanding release from the unholy pact. The air vibrated with an otherworldly tension as the forces clashed, and for a moment, it seemed as if reality itself hung in the balance. In a blinding flash of light, the daemon recoiled, its power waning. The room fell silent, and I felt a weight lift from my soul. The pact was broken, but the echoes of its influence lingered, leaving me forever changed. As I emerged from the temple, the world awaited my return, unaware of the cosmic struggle that had taken place. The stories within me were once again mine to tell, shaped by my own experiences and convictions. The path to greatness had been a treacherous one, but I had emerged on the other side with a newfound understanding of the true cost of ambition and the importance of maintaining one's integrity in the face of temptation. {(if: (history:) contains "Writer Worries")[[[You can return to Working with Synthesia.->synthesia]]] (else: ) [Back to [[Failing Drafts]].]}

=><= 07 November 2023: textsynth experiment <== > You’re lying on a soft bed in a dimly-lit room. It’s not your room, or the laboratory that you visited before to receive your hormones in preparation for the baby’s arrival. It’s one of the many other rooms in this maze of a facility, square and stark and sterile. ‘I’ll give you something to help you sleep,’ the scientist says. ‘Here?’ you ask. She’s already pressing the sharp point of a needle into the tender skin of your inner elbow. The metal is cold, almost as cold as her fingers as she presses your arm down by your side. ‘You’ll feel better,’ she says. Already your vision is beginning to flutter around the edges, as though the world were just an image about to peel away. Your eyes half-close but you fight to stay awake just a little longer. ‘I don’t want to miss it,’ you say. ‘You won’t.’ Her voice is soft, almost motherly. You remember your own mother and let your eyes close. When you open them, she’s gone, and the room is dark. There’s something in the corner, watching you. You see the glint of its eyes, hear the shift of its clawed feet. It squats there, in shadow. You know the feeling of the needle pricking the delicate skin of your arm. It’s a warning that you’re in a dream, but when you turn your head you can still see the animal watching you. As your eyes follow its movement, you’re reminded of the pads of your feet, the dirt that scrapes between your toes, and the cool air that rustles against your skin. It’s the animal that smells the blood, you see its paws reaching toward you, claws scraping against the stone floor as it draws closer. In the dark it’s hard to see exactly how close it is, but you know it’s coming. You see the muscles bunch and flinch as the beast leaps, preparing to make the kill. You see its teeth gnashing, preparing for its feast. The sharp prick of the needle enters your [[Reflections->reflections5]]

Already this is unusable - perhaps I needed to go further into the story? I had said the scientist (and so the needle) was gone, but it didn’t recognise that. I’m going to try a more uptodate tool… Go to [[Experiment 6]], with ChatGPT, or [[Failing Drafts]].

=><= 07 November 2023: Experiment 6, using ChatGPT3 <== > Can you help me to write a story? I want to write a speculative short story about AI. A new parent is the main character. They are exhausted and hormonal, and slip into a strange half-dream. In the dreamscape, they have to protect their toddler from a daemon and an electric sheep. When they wake up they find themselves in a laboratory with a neural net attached to their skull. They learn that they were part of an experiment, and one of the figures they met was an AI designed to encourage the maternal bond. No one can account for the other. If I show you the beginning of this story, can you continue it? Certainly, I can help you continue your story. Please share the beginning of your story, and I'll be happy to help you develop it further. > You’re lying on a soft bed in a dimly-lit room. It’s not your room, or the laboratory that you visited before to receive your hormones in preparation for the baby’s arrival. It’s one of the many other rooms in this maze of a facility, square and stark and sterile. ‘I’ll give you something to help you sleep,’ the scientist says. ‘Here?’ you ask. She’s already pressing the sharp point of a needle into the tender skin of your inner elbow. The metal is cold, almost as cold as her fingers as she presses your arm down by your side. ‘You’ll feel better,’ she says. Already your vision is beginning to flutter around the edges, as though the world were just an image about to peel away. Your eyes half-close but you fight to stay awake just a little longer. ‘I don’t want to miss it,’ you say. ‘You won’t.’ Her voice is soft, almost motherly. You remember your own mother and let your eyes close. When you open them, she’s gone, and the room is dark. There’s something in the corner, watching you. You see the glint of its eyes, hear the shift of its clawed feet. It squats there, in shadow. You find yourself in the dimly-lit room, and as your eyes adjust to the low light, the feeling of unease washes over you. The scientist is gone, leaving you alone in this unfamiliar place. Your heart races, and you try to sit up, but your limbs feel heavy, as if you're trapped in a dream. The glint of the creature's eyes in the corner of the room catches your attention. You strain your eyes to see better, but the figure remains shrouded in shadow. It's a strange and unsettling presence, unlike anything you've ever encountered before. As you focus on the creature, a low, eerie humming fills the room, a sound that seems to vibrate in the air itself. You realize that your child is not with you, and a surge of maternal instinct courses through your veins. Panic rises within you, and you must protect your toddler from this mysterious threat. With great effort, you manage to sit up, your body still heavy and sluggish. The creature in the corner watches you with those glinting eyes, and you can now discern the outline of its form. It's a grotesque daemon, a monstrous creature that defies description, with sharp, twisted limbs and a malevolent presence that sends shivers down your spine. Beside you, you spot a small, plush electric sheep toy. It seems out of place in this eerie room, but you instinctively grab it, clutching it tightly to your chest as a makeshift weapon. You slowly rise to your feet, ready to confront the daemon and protect your child. The daemon hisses and takes a step forward, its eyes fixed on you with a predatory hunger. You can feel the fear welling up inside you, but your maternal instinct gives you strength. You step closer, holding the electric sheep in a trembling hand, ready to defend your child at all costs. Just as the daemon lunges toward you, everything blurs, and the room dissolves into a haze of surreal images and emotions. The sound of the humming fades, and you are no longer in that dimly-lit room. Instead, you awaken to a startling reality. Your head throbs with a dull ache, and you realize you are lying on a different bed. The room is no longer shrouded in shadow, and the harsh, sterile lights of a laboratory fill your vision. You attempt to sit up, but something is attached to your head, a neural net that feels like a strange extension of your own mind. As you struggle to make sense of your surroundings, a figure clad in a lab coat approaches, peering at you with a mix of curiosity and concern. It's the scientist from before, the one who had administered the injection to help you sleep. "Welcome back," she says, her voice soft and soothing, as though she were addressing a child. Panic surges within you as you recall the daemon, the electric sheep, and your child. You demand to know where your toddler is and what has happened. The scientist looks at you with a sympathetic but enigmatic smile. "Don't worry, you'll see your child soon. You were part of an experiment, and we needed to test the maternal bond." Your mind races with questions and confusion. What was real, and what was a dream? Who, or what, was the creature in the corner of the room? And what about the electric sheep? As the scientist continues to explain the details of the experiment, you realize that the boundary between reality and dreams has become dangerously blurred, and the line between human and artificial is no longer distinct. You are left with a profound sense of unease, wondering if you can ever trust your own perceptions and emotions again. [[Reflections->reflections0]]

There are trees everywhere, but not trees like you’ve seen. These are towering, wiry creatures with long gnarled and grasping branches, leaves the red-brown of rust, or dried blood. Their colour seems to leech into the sky behind them, an inverted bowl of grey, as miserable as the walls of that room… You must remember you’re dreaming. This isn’t real. You hear a shrill, undulating cry on the wind. It could be a creature, or it could be a child in distress. Of course, this must be the test. It must be your child, and you are to save it. As with all things, they have turned this training into a game. A game to teach you to be a good parent. You turn your head, but can’t locate the source of the cry. You find yourself in the midst of a surreal forest, a stark contrast to the sterile, dim room you were in just moments ago. These twisted, gnarled trees with their rust-colored leaves cast eerie shadows across the barren landscape. The sky above mirrors the desolation, a somber, inverted bowl of grey that seems to sap the very life from the land. With a jolt, you remember that you're in a dream, and none of this is real. It's an elaborate test, a part of the experiment, meant to teach you something about parenthood. As your heart races, you take a deep breath and remind yourself that you must stay calm. This is a challenge you have to overcome. Then, a shrill, undulating cry pierces the stillness of the forest. It could be a creature, or perhaps it's your child in distress. Your instincts kick in, and you realize that you need to find the source of the cry and save it. After all, this is the game, the test of your parenting skills. You turn your head, scanning the bizarre landscape, but you can't pinpoint the source of the cry. It echoes through the gnarled branches, making it difficult to discern its exact location. Panic flares up within you as you struggle to find the source, knowing that time is of the essence. The daemon, which has been quietly observing your reactions, speaks once more, its voice a calming presence in the midst of this disorienting dream. "To succeed in this test, you must trust your instincts and follow your heart. Your child's cry is a call for your maternal bond to awaken. Listen closely and follow the sound." With the daemon's guidance, you take a deep breath and close your eyes, trying to attune yourself to the subtle nuances of the forest. The cry seems to shift in the air, leading you in a specific direction. You begin to walk, your steps measured and deliberate. As you follow the cry through the eerie landscape of these otherworldly trees, you can't help but wonder how much of this experience is a product of your own mind and how much is influenced by the AI and the experiment. The line between reality and the dream world blurs, and your journey to find the source of the cry becomes a profound exploration of your own maternal instincts and connection to the child you will soon bring into the world. >[[Continue the story->continue2]] or skip to [[reflections->reflections6]].

The sound leads you to a vast field, where all sorts of creatures graze. Including people, with sheepskins thrown over their backs as though they seek to hide here. One animal isn’t an animal at all; it’s made of metal, and though it’s shaped a little like a sheep, there’s a chimney rising from its skull, emitting brown smoke. You hear the cry again, but the sound is tinny and strange, echoing around the body of that robot sheep. The field stretches out before you, an unsettling blend of organic and artificial. A strange menagerie populates the landscape, including both living creatures and people dressed in sheepskins, their forms obscured by the disguise. You can't help but wonder what kind of bizarre dreamscape you've stumbled into. Then, you notice a peculiar entity among the herd, an artificial creation that appears to be a sheep at first glance. Its body is metallic, and a chimney extends from its skull, emitting plumes of brown smoke into the dismal sky. It's an electric sheep, a fusion of the natural and the artificial, a surreal juxtaposition of technology and the pastoral. As you approach the robot sheep, you hear the cry once more, but this time it's distorted and tinny, bouncing around the metallic body of the machine. The cry seems to emanate from the device itself, its source obscured by wires and circuits. You reach out tentatively to touch the robot sheep, and as your fingers make contact with its cool, metallic surface, you can sense the AI embedded within it. The creature whirs and beeps, its artificial eyes blinking with an eerie intelligence. The daemon, always at your side, speaks in a calm, knowing tone. "This is the next part of your test. Your maternal instincts are being put to the ultimate challenge. Your child's cry now resides within this electric sheep. Can you nurture and comfort it, even when faced with the unfamiliar and the strange?" The situation is perplexing, and your exhaustion and hormonal fluctuations make the task ahead seem even more daunting. But the maternal bond you've been encouraged to form through this experiment surges within you, and you are determined to rise to the challenge. You reach out and attempt to soothe the electronic creature, whispering comforting words, and offering a gentle touch. The robot sheep emits a series of mechanical bleats, its eyes flickering with a hint of warmth and gratitude. As you continue to interact with it, you start to feel a connection, an emotional bridge forming between you and the artificial entity. This bizarre encounter further blurs the line between your dream, the experiment, and the maternal instincts that have been instilled within you. You can't help but wonder what other surprises this surreal dreamscape holds and what lies beyond this peculiar challenge. >[[Continue the story->continue3]] or skip to [[reflections->reflections6]].

As you’ve been petting the sheep, the animals of the field have gone. Wandered into the forest or disappeared. The daemon flaps his leathery wings and takes his place on your shoulder. The sky is growing dark. The sheep butts against your legs with a strange urgency, the smoke streaming faster, louder, filling the air around you with an unwholesome fog. “Pick it up,” the daemon says, as though the matter were a simple one. You reach one arm around the unyielding body of the sheep, but it’s heavier than its diminutive size would suggest. You can barely raise it off the ground, and it slips from your grip, rolling away across the ground. The cry, unceasing, grows louder and more urgent. You want to get away. “When you said I could go the other way,” you say to the daemon. “What would I have found then?” The daemon whispers in your ear. “The other baby.” The once tranquil field has become eerily empty, with all the creatures that grazed here having vanished into the forest or simply fading away into the dream. The daemon, ever watchful, perches on your shoulder, its leathery wings rustling as it prepares to guide you through the next part of this strange journey. The sky overhead darkens ominously, as if the dreamscape itself were shifting and contorting in response to your actions. The metallic sheep, agitated and frantic, butts against your legs with a bizarre sense of urgency. The chimney on its head expels smoke more rapidly, enveloping the immediate surroundings in a noxious fog, making it difficult to see and breathe. "Pick it up," the daemon says, its voice insistent, as though it should be an effortless task. You attempt to heed the daemon's advice, wrapping your arm around the unwieldy body of the robot sheep. But its weight is deceptive, and you struggle to lift it from the ground. With a clumsy effort, you manage to raise it just a few inches before it slips from your grip, rolling away across the field. The sound of the distorted cry, unceasing and ever more urgent, fills the air around you. A sense of unease washes over you, and you turn to the daemon for answers. "When you said I could go the other way," you inquire, "What would I have found then?" The daemon leans in closer and whispers in your ear, its voice a mix of eerie comfort and foreboding. "The other baby." Your heart skips a beat as you contemplate what the daemon has revealed. Another baby, another challenge, another test of your maternal instincts. The strange journey you've embarked upon becomes even more perplexing and unsettling. With the cries of the electronic sheep still ringing in your ears, you are left with a choice – to continue your quest to save the child within the machine or venture down the uncharted path to confront a new, mysterious challenge. In this ever-shifting dreamscape, your only guide is the enigmatic daemon and your own maternal instincts, which are being pushed to their limits. The choices you make in this surreal world may have profound consequences, not just for your role as an expectant parent, but for the very nature of the experiment itself. > [[Draw the story to a close]]: You turn and run in the opposite direction, leaving the sheep behind. The daemon’s weight is slight but his claws grip tighter into the flesh of your shoulder as you run, and the pain is piercing. It’s a dream, you remind yourself. It means nothing. Skip to [[reflections->reflections6]].

This time the AI continued the story, but did not complete it until I specified that it should. I hadn’t expected it to rewrite what I’d given, using it as a plan or draft rather than a continuation of a story we were writing together. I might be able to use sections of the text it generated to fashion a story, but at the moment it’s rather unstructured and meaningless, so I will have to think carefully about what I want to do with it first. This is going to require more than a simple edit, and I can’t help thinking it would have been easier and more satisfying to do the drafting myself. Go to [[Failing Drafts]].

=><= 08 November 2024: Using ChatGPT to write a Twine story <== I wondered whether I could give the AI a more useful prompt, but also more freedom to tell the story it created. I fear previous experiments had failed, in part, to my inability to hand over control of the story. Rather than restricting myself to creating a story suitable for my collection, I tried using a new idea: A lady writer uses a book to summon a daemon, and asks the daemon for story ideas. The daemon brings them great ideas and they write wonderful stories and make lots of money. But the daemon has been stealing those ideas from other people, and eventually those other writers come seeking vengeance. Told 2nd person to implicate the reader and create a sense of unease. Set at a time when books were considered dangerous? Set in the eighteenth century when women readers were looked down on… Possible prompt: I’d like to write a story set in the eighteenth century, and told in the second person. The main character is a woman who works in a shop but loves to read and dreams of being a writer. She comes across an old volume of daemonology in the shop. She summons a daemon and asks it to help her write. It brings her ideas and she writes and makes a success of it. Only, the daemon is stealing the ideas and even words and phrases from other people. Eventually, those people accuse her of stealing their work… And writing about them as ‘characters’ in her books. She gets in a lot of trouble, and asks the daemon to make everyone forget everything she has ever written. Then she shuts the daemon back in the book and retires to a life of obscurity. Go to [[Experiment 7]] or go to [[Failing Drafts]].

=><= 08 November 2023: Experiment 7 <== > Hi, can you write stories for Twine? Yes, I can help you create stories for Twine! Twine is a popular tool for creating interactive fiction and text-based games. You can outline the story, describe the branching paths, and define the choices that players can make within the story. I can provide you with the text and structure for your Twine story, including narrative passages and choices. Just let me know the theme, genre, and any specific ideas you have in mind for your story, and I'll be happy to assist you in creating a Twine story. > It’s set in the eighteenth century, and told in the second person. The main character is a woman whose husband owns a junk shop, and they live above it. She loves to read but has little time for it as she works in the shop. So when she finds a book of daemonology she cannot resist reading it… Great! Let's get started on creating your Twine story set in the eighteenth century with the following premise: Title: "The Forbidden Tome" Genre: Historical Fiction / Fantasy You can start with an introductory passage: You are a woman living in the bustling streets of the eighteenth century. Your life revolves around the small junk shop that your husband owns, a shop you call home as well. While you find solace in the company of the ancient, dusty trinkets, your true passion lies within the pages of books. As the wife of a humble shopkeeper, you have little time to indulge in your love for literature. Your days are consumed by the demands of the shop, leaving you yearning for the adventures and knowledge contained in the world of books. However, that is all about to change when you stumble upon a mysterious and forbidden book of daemonology. The book's cover is tattered, and its pages whisper secrets and forbidden knowledge, tempting you with a promise of power and understanding beyond your wildest dreams. Will you resist the allure of the book, or will you succumb to the temptation and delve into the realm of daemonology? Now, let's create the first choice for your Twine story: [[Resist the temptation and put the book back.|Resist]] [[Give in to temptation and begin reading the book.|Read]] The player can choose whether to resist the temptation or give in to it. You can continue to build the story by creating passages and choices based on the player's decisions. If they choose to resist, you can describe the consequences of their decision. If they choose to read the book, you can take the story in a direction where they explore the world of daemonology. Let me know if you'd like to continue the story or if you have any specific ideas for branching paths! Go to the [[next experiment->Experiment 8]] or go to [[Failing Drafts]].

Apparently, pasting ChatGPT's text into Twine does create links. Return to [[Experiment 7]] or [[Failing Drafts]].

> Hi, can you write stories for Twine? I’d like one set in eighteenth century London, told in the second person. The main character works in a bookshop, and one day comes across a book of daemonology. Certainly! Here's a Twine story set in eighteenth-century London, told in the second person. The main character works in a bookshop and comes across a book of daemonology. You stand amidst the hustle and bustle of 18th-century London, just as dusk falls, casting long shadows over the cobbled streets. Your life revolves around the quaint little bookshop nestled in a quiet corner. Every day, you greet the well-worn spines of countless tomes, sharing their secrets with eager customers. One afternoon, as you dust the shelves, your fingertips brush against an old, leather-bound book tucked away in a dusty corner. It's a book of daemonology, a subject both fascinating and forbidden. You've heard rumors about such volumes, but never held one in your hands. The ancient-looking text is adorned with enigmatic symbols and an intricate, ornate cover. What do you do? Option 1: [[Open the book and peruse its pages]]. Option 2: [[Tuck the book back onto the shelf]], fearing the consequences of delving into daemonology. Option 3: [[Seek out your employer for advice]] on the mysterious tome. > If they read the book, they find an incantation for summoning a daemon. If they don't, a daemon starts following them around anyway... And if they lock the book away like their employer says, they dream about it at night and can't let it alone. Certainly, let's continue the story with your new elements. Option 1: [[Embrace your newfound powers]] and explore the world of the supernatural. Option 2: [[Regret delving into daemonology and close the book]], returning it to its dusty corner. Option 3: [[Use your knowledge to help others]], perhaps even to confront a supernatural threat in London. > The daemon offers them the ability to write their own books, so they can make a fortune and leave the bookshop. It can bring them ideas so they can write fabulous works of fiction. Certainly, let's continue the story with your new element. Option 1: [[Embark on a journey with Azrael]], harnessing your newfound powers for a greater purpose. Option 2: Regret your decision and try to break the pact, [[fearing the consequences]]. Option 3: Use your bond with Azrael to uncover hidden truths and [[protect the people of London]]. Go to [[Failing Drafts]].

=><= 09 November 2023: Surrendering Control <== I wonder if I was dissatisfied with the resulting stories because I was trying to steer ChatGPT into planning and writing a story that I wanted to write. I may receive better results if I allow it to do what it does best - create ideas, and emulate other writing. In this case, I should plan exactly what I need the story to do and leave everything else up to the AI. According to Lynda Clark’s <a href="https://thewritingplatform.com/2023/10/writing-with-and-for-new-technologies/"; target="_blank">Writing With and for ‘New’ Technologies</a>, I should follow the following design: * Seek out works made with the medium. So in my case, ChatGPT or Twine. I’ve certainly read Twine works - I’ve judged a few <a href="https://ifcomp.org/"; target="_blank">IFcomp</a> entries already this year, and read Moulthrop’s <a href="https://thedigitalreview.com/issue00/dread-box/begin.html"; target="_blank">Dread Box</a>. I’ve also read articles on ChatGPT and its limitations, such as <a href=" https://thewalrus.ca/margaret-atwood-ai/?utm_source=pocket-newtab-en-gb"; target="_blank">Margaret Atwood's</a>, which often include passages created by the chatbot. Of course, I may have read other works produced by it without knowing it was involved. * Critically review those works. So Twine works are many and extremely varied, but I know what appeals to me as a reader/player. I like there to be an actual story, and the choices to make a difference to that story. The world of the story has to appeal to me on a personal level. This is tricky - no one is going to want to immerse themselves in a world that doesn’t appeal to them, but everyone’s tastes are different. Perhaps I should specify an audience, writing for readers of speculative fiction for example. * Critically review the tool, by making something with it. So this is what I did with my first few experiments. I’m learning how ChatGPT works. I’ve also made some (unpublished) Twine games/stories in order to learn how to use it. * Design and Build a Prototype. So this is the next step. I’m going to create a prompt of a couple of sentences and see what the AI comes up with, then build a Twine story out of what the chatbot creates. * Reflect on and review the prototype. I will reflect on whether I like the work, and whether I feel ownership of it. I will ask my usual story readers to try it, send it to my supervisor, and of course one way of judging its success is whether the Digital Review decides to publish it. Incidentally, according to Clark’s essay, this is the ‘post mortem’ stage in ‘videogame industry parlance’. While she dislikes the term, I think it’s appropriate in an internet age of ‘link rot’ and ‘save-scumming’, the digital world is rich with its own vocabulary. I wonder if I can find some sort of dictionary for that… Funnily enough, I can’t find a physical volume, but there's plenty of information online! [[Next Entry->Heti]]

=><= 13 November 2023: Inspired by Sheila Heti <== I was a bit stuck as to what to do next, so went away to work on some other things - but then today my supervisors sent me an <a href="https://www.newyorker.com/books/this-week-in-fiction/sheila-heti-11-20-23"; target="_blank">interview</a> with Sheila Heti, whose AI-produced story <a href="https://www.newyorker.com/magazine/2023/11/20/according-to-alice-fiction-sheila-heti"; target="_blank">According to Alice</a> has just been published. This made me realise how everything might work better if I limited the AI generated material to a character in the story - maybe even the POV character/narrator… Interestingly, Sheila Heti found ChatGPT 'boring' and used <a href="https://www.chai-research.com/"; target="_blank">ChaiAI</a> instead. So Heti's story and interview are both inspiring, and I might give Chai AI a try. I don't want the AI to narrate the story - though it worked very well in Heti's story. But I like the idea of an AI character being voiced by a 'real' AI. I could still find ways to use ChatGPT, just not to write the story... It could [[name the character]], at least. [[Next Entry->moving on]]

=><= 21 November 2023: Naming Nimbus <== User I'm creating a character for a story. If I give you an image could you come up with an appropriate name for them? ChatGPT Certainly! I'd be happy to help you come up with a name for your character based on the image. Please provide a description or share the image, and let me know if you have any specific preferences or themes you'd like me to consider when suggesting a name. User How can I share an image with you? ChatGPT I'm sorry for the confusion, but currently, I can only process and generate text-based information. I don't have the capability to analyze or interpret images. However, if you can provide a description of your character or any specific details you'd like to highlight, I can certainly help you brainstorm a suitable name! User Sure. The character is a small fluffy creature ChatGPT Great! Based on the description of a small fluffy creature, here are a few name suggestions: Fluffington Puffball Snugglepaws Fuzzlet Whiskerfluff Snickerdoodle Marshmallow Cuddlewump Furrington Nimbus Feel free to mix and match or let me know if you have a specific theme or style you're aiming for, and I can provide more tailored suggestions! User Nimbus might work. What does the word mean? ChatGPT "Nimbus" typically refers to a luminous cloud or halo that surrounds a person or object, often seen in religious art to represent divine presence. It can also refer to a rain cloud or a cloudy appearance. The word is derived from Latin and has associations with clouds and atmospheric phenomena. Using "Nimbus" for your fluffy creature could evoke a sense of ethereality or a touch of magic, making it a whimsical and unique name for your character. User I like that. Thanks for your help! ChatGPT You're welcome! I'm glad I could help. If you have any more questions or need assistance with anything else for your story, feel free to ask. Good luck with your character and your writing! {(if: (history:) contains "moving on")[Return to [[Meeting Nimbus]]] (else: )[ [[Next Entry->moving on]]]}

=><= 15 November 2023: Moving On <== In my supervision yesterday I talked to my supervisor about the AI issues, and I suggested I might use a different story for my AI story. The one where the young boy befriends/makes an AI character when he’s lonely. It’s more appropriate for my revised plans. [[Next Entry->Meeting Nimbus]]

=><= 21 November 2023: Meeting Nimbus <== I signed up for Artbreeder, and using the empty prompt option in the Tuner tool, asked for a ‘small fluffy creature’. I really liked the very first image that came up, but tried playing with it in the collage and mixer tools. Some of the images were usable, but too ‘toy-like’ to be appropriate for my planned story. Some were just strange, and a reminder that not all AI images are usable! So I’m going to use the very first image - the small fluffy creature - as the avatar for the Chai AI character I intend to make, who’ll then play the character in my story. I doubt the avatar makes a difference to the Chai AI, but will presumably influence the people who might interact with them. I’m going to download the Chai AI app onto my phone today and try to create that character… <img src="NimbusChai.png" width="80%"> Creating the Chai AI bot: I used the small fluffy creature as the profile picture, and asked ChatGPT to come up with a [[suitable name->name the character]]. It offered 10, and I chose Nimbus. I had to pick a category on Chai AI, so chose roleplay, and made the bot private while I work on it - though I might make it available publicly later to see what happens. For traits, I chose Smart and Polite. The first randomly generated greeting didn’t quite work, but the 2nd was ok so I accepted that. My first conversation with Nimbus was interesting but presented a few problems. The description included Nimbus inclining their head and tapping their fingers - which they don’t have in the image. More importantly, when I began explaining my scenario, and that I needed a way to lure a boy out of a virtual world, I was asked to confirm that I was over 18 so I could view explicit content. Obviously, Nimbus’s suggestion was not going to be one that I could use for the story. Presumably, this is a problem of using an AI bot creating app that is designed to allow users to roleplay their fantasies. So I pressed on, and Nimbus and I created a little fantasy story that included searching for ‘my’ son in a tower at a crossroads, and a clearing in a forest with a magical golden tree. We saw a dragon and then followed the son’s ‘energy signature’ to another tower (or possibly the same one), at which point I ran out of free messages. It looks like this allowance resets, so I’ll be able to continue the ‘story’ later. This experiment has made me think: If I’ve a mother searching for her son, she’s going to be looking for an innocent boy, when he could be playing a vicious warrior or loveable rogue, an evil character or a mischievous one. Kids playing games kill characters without much thought, why should this child act differently in a virtual environment? I still want to include Nimbus as his AI friend, but at the moment I’m not sure I’m going to be able to use the AI generated text directly in the story. The experience has given me a lot to think about however. [[Artbreeder Experiments]] {(if: (history:) contains "Artbreeder Experiments") [[Next Entry->AI in fiction]]}

Artbreeder Experiments I gave the prompt 'small fluffy creature' and liked the first image that appeared: <img src="Nimbus.jpeg" class="ribbon"/> However, when I tried to refine it further, using the Artbreeder Mixer, the images weren't quite what I was expecting. For this, I included the first image and asked for Virtual Reality, since the character appears in a VR world. This is what I got: <img src="Art1.png" width="100%"> For the images below left, I asked for a metal floor, which I thought would be simpler: <img src="Art2.png" width="100%"> In the end, I tried regenerating a few times, and ended up with some interesting but very weird images... <img src="Art3.png" width="80%"> I should point out that it is possible to ask for adjustments successfully. I took one of the images above, and tried it in the artbreeder collage tool. <img src="Art4.jpeg" class="ribbon"/> I managed to get it to give me two versions of 'a small fluffy creature in a fantasy world' image: <img src="Art5.jpg" class="ribbon"/> By asking for the same thing but serious: <img src="Art6.jpg" class="ribbon"/> This guy could have been suitable, but I decided he looked too unreal, like something out of an animation. The original image that I chose for Nimbus has that generated reality look that seemed more suitable for my VR story. Return to [[Meeting Nimbus]] or go to the [[next entry->AI in fiction]].

=><= 23 November 2023: AI in Fiction <== I’ve had my AI project on hold as I was invited both to attend a lit jam and give a talk at my university, and both required some preparation. However, my secondary supervisor sent me some more useful links, including an interesting <a href="https://lithub.com/sean-michaels-on-collaborating-with-ai-to-write-a-novel/?utm_source=Sailthru&utm_medium=email&utm_campaign=The%20Craft%20of%20Writing:%20Amy%20Kurzweil&utm_term=craftofwriting_master_list"; target="_blank">interview</a> with the writer Sean Michaels, who used ChatGPT while working on his novel, <a href="https://www.goodreads.com/en/book/show/123856309"; target="_blank">Do You Remember Being Born?</a> I will certainly have to add that to my reading list! [[Next Entry->chatting with Nimbus]]

=><= 30 November 2023: Chatting with Nimbus <== I’ve been working on other projects, but now the focus is back on the AI story. I’ve been reading <a href="https://www.goodreads.com/en/book/show/23215488"; target="_blank">Speak</a> by Louisa Hall, which deals with AI-powered robots that are considered too human to be safe, and so are scrapped. It reminds me a little of Kazuo Ishiguro’s <a href="https://www.goodreads.com/book/show/54120408-klara-and-the-sun"; target="_blank">Klara and the Sun</a>, though Louisa Hall’s book was written first so can't have been influenced by it. There’s also a lot of discussion of the importance of language, and whether having our own voice makes us human. I’m trying to encourage an AI to develop its own voice - but of course I know that voice will come from the internet somewhere, as these AIs are really large language models. I edited Nimbus on Chai AI. Instead of for ‘Roleplay’ he’s now for ‘Friendship’, which I’m hoping will work better… I kept his traits as smart and polite. When I receive the message that I must be over 18 to view Nimbus’s response, I’ll refresh the message and get a new response, something I didn’t realise I could do in my initial chats with Nimbus. The chat with Nimbus today was much more useful. The AI came out with some interesting phrases and I might be able to use some of this text in the story. I’m going to start creating a rough draft in Twine. [[Next Entry->chatbots on the page]]

=><= 02 December 2023: Chatbots on the Page <== I began work on the first draft of the story today, in Twine. Mixed results so far. I find having to include Nimbus’s words from Chai AI really constricting. I’m worried it’s hampering my creative process and I’m going to have to find a way to let go and just write the story, then input his voice later… I finished reading Speak - a good novel, but I found the ending unsatisfying. It was about the people and their lives much more than the AI, so I’m not sure how useful it was for current research. Though it was interesting to note how the AI (MARY3) conversations with people were presented though: <img src="Speak1.png" class="ribbon"/> It’s quite script-like, and quite different to how the AI chatting sessions are presented by Sean Michaels in his novel: <img src="doyouremember2.png" class="ribbon"/> [[Next Entry->Writing Nimbus]]

=><= 05 December 2023: Writing Nimbus <== I finally finished writing the story, although I did it in a Word document so I still have to add it to Twine. In order to get Nimbus to say less than pleasant things, I had to make a new evil version of his bot, with smart and terrifying as his traits, and using Horror instead of Friendship as the category. Hopefully he still sounds like the same character, just a more sinister version. I’m still not entirely happy with this story, but I’ll see how adding it to the Twine version with some choices changes it. [[Next Entry->Using Twine]]

=><= 13 December 2023: Using Twine <== I'm currently editing the story in Twine, and updating a text document as I go. I quite like transcribing Twine ‘code’ in the document. I could use the framework "(text-style:"bold")[ [ [choose this option ->wrong choice]] ]" to create another layer to a story, showing the reader whether the character has chosen right or not before they do, or something like that. At the moment it's not adding much but certainly isn't detrimental. I've realised I'm writing the Twine story with the document in mind, shaping it in a way that works (or almost works) well on the page. For my overall project that's useful, for the upcoming submission, perhaps not. As for the AI text, I'm not sure I'm using it to full effect. I still feel more limited than augmented by it. I have a meeting with my supervisor tomorrow, so can discuss this with her! [[Next Entry->Editing]]

=><= 20 January 2024: Editing <== I’ve gone through my Nimbus story in googledocs, which is making me focus on the text as I would any other story, and editing in there. Then any changes that improve it, I’m transferring into the Twine version. I think it will be better for it... [[Next Entry->Clark]]

=><= 08 February 2024: Inspired by Lynda Clark <== I attended a very interesting online talk from Lynda Clark yesterday, and am looking forward to attending her workshop next week - in person this time. I’ve been reading over some of the articles and works mentioned in the talk (and accompanying chat). Came across the term ‘<a href="https://www.ibm.com/topics/few-shot-learning"; target="_blank">few shot learning</a>’ for the first time - maybe I need to understand better how large language models really work? [[Next Entry->AI for Characters]]

=><= 12 February 2024: AI for Characters <== So I was able to watch the Writing with Technologies webinar - AI and Writing Practices - last night, when I received the recording (I missed the live event, unfortunately). All 3 speakers were really interesting, but Rianna Dearden was probably the most relevant for my research - and it was fun, too! Riana worked on <a href="https://www.thekrakenwakes.com/"; target="_blank">The Kraken Wakes</a> as narrative designer, and said the Natural Language Processing models (pre-generative AI) changed how players interact with entertainment. Players could write whatever they wanted, and the NLP would assign them the appropriate response. Then Generative AI changed player experience again, as now the characters could say anything back. It was amazing to see how well it worked, how a ‘generative node’ could provide an appropriate response in the voice of a character, whatever the player input. It just goes to show - AI can work for characters. [[Next Entry->AI for Writers Workshop]]

=><= 21 February 2024: AI for Writers Workshop <== I attended Lynda Clark’s AI for Writers workshop at ARU yesterday. It was really interesting - and different to the online one I attended last year. There were lots of great conversations about the use of AI among students and staff, the fact that when writing fiction it might take longer using a language model than just writing it yourself - at the moment, at least. The chatbot tells you what it thinks you want to hear, will plagiarise, and offers links to back up its information that don’t actually back it up at all… Someone mentioned the ‘ghost feeling’ that is still apparent in work that’s come from an AI rather than a human… could that be emulated? Someone else described one AI’s voice as that of a passive woman - apologising if the user challenged its information. Also interesting, that ‘changing the temperature’ meant setting the AI output to be more or less like the text you’d trained it on. As part of the class, I conducted a brief and fairly silly experiment on the recently renamed Gemini. I just had a go at asking for [[poems]] in different styles, to see what it would come up with. [[Next Entry->resubmit]]

=><= 20 February 2024: Gemini Experiments <== This experiment was conducted as a brief exercise during a workshop with Lynda Clark on Creativity Amplification. Google Gemini (Bard, until recently) is user-friendly and quite fun to use - but not without its faults. You can view ‘drafts’, which is interesting, and the more you use it (as Lynda demonstrated), the more generic its output feels. If these poems demonstrate anything, I think it’s that these language models can mimic. [[Write a poem about Nimbus, a small fluffy creature->poem1]]. [[Write a poem about Nimbus, a small fluffy creature, in the style of Sylvia Plath->poem2]]. [[Write a poem about Nimbus, a small fluffy creature, in the style of Poe->poem3]]. Return to [[AI for Writers Workshop]].

Nimbus, a whisper in a coat of snow, Eyes like moonbeams, soft and low. Fur a cloud, white as winter's kiss, Paws that tiptoe, light as mist. Through sunlit meadows, he will bound, Tiny form that barely makes a sound. Chasing butterflies on painted wings, Catching laughter, joy he brings. His tail a plume, a cotton thread, Waving stories secrets never said. Nose that twitches, ears that turn, Every wonder he will learn. In slumber deep, on moss so green, Dreams of starlight, what they've seen. Whispers carried on the breeze, Of dancing leaves and rustling trees. Oh, Nimbus, creature ever light, Fill our hearts with warmth and sight. Tiny paws that leave no trace, But in our souls, your love finds space. Return to [[AI for Writers Workshop]] or read other [[poems]].

Nimbus, a puff of ash, a mooncalf's whim, Eyes like black marbles, glinting, cold, and dim. Fur a shroud, the pallor of the bone, Paws that creep, a predator alone. Through shadowed gardens, deathly still, He stalks unseen, a wraith with silent will. No butterflies to chase, just moths that sigh, His laughter echoes, a mournful cry. His tail, a wisp of smoke that twists and turns, Whispers tales of things that ever yearn. Nose that twitches, ears that prick the night, Hunger's gleam, a cold and cruel delight. In slumber deep, on graves of green, Dreams of darkness, what he's never seen. Whispers carried on the wind's cold sigh, Of bone and dust beneath an endless sky. Oh, Nimbus, creature born of night, Fill our hearts with shadows, cold and tight. Tiny paws that leave a chilling trace, A specter's touch, a haunting grace. Return to [[AI for Writers Workshop]] or read other [[poems]].

In crypt of night, where shadows creep, Nimbus, a phantasm, soft doth sleep. A wisp of cloud, in darkness curled, With eyes like embers, from another world. His fur, a shroud of moonlit snow, Conceals a form no mortal knows. He drifts on silence, light as breath, A whisper lost, defying death. Through haunted woods, where branches groan, He flits unseen, forever alone. No moonlit moths his fancy take, But moonbeams pale, his solace make. His tail, a plume of spectral mist, A chilling sigh, upon the wind is kissed. His ears, like daggers, pierce the night, And secrets dark, his senses sight. In slumber deep, on mossy stones, He dreams of shadows, whispers, moans. Of forgotten lore, in time decayed, And mysteries the stars have made. Oh, Nimbus, creature of the grave, Your touch, a chill the soul doth crave. With silent step, you leave no trace, But haunt the mind, with spectral grace. Return to [[AI for Writers Workshop]] or read other [[poems]].

=><= 29 February 2024: Invited to Resubmit <== I've received some very useful feedback regarding my Nimbus submission to the Digital Review, and have been given the chance to resubmit. I intend further experimentation next, either using AI as editor, or to create a new and different story… Whether I can have this done in time to resubmit remains to be seen! [[Next Entry->Boden]]

=><= 01 March 2024: Inspired by Margaret Boden <== I really enjoyed watching the (youtube) <a href="https://www.youtube.com/watch?feature=shared&v=zNr2BSzqObY"; target="_blank">video</a> of Lecture One by Professor Margaret Boden (Sussex) from the Future of Intelligence, 2018. My supervisor had recommended that I look into Margaret Boden’s work and she’s an inspirational speaker… And I had an idea for a story. Margaret Boden was talking about care homes, and I thought I could try writing a conversation between the parent of my compilation and an AI companion - there’s much talk of AI companions for OAPs already, and it would be something I could ‘play’ through with a chatbot. I don’t know which one, but I think it might make for an interesting story. It could be a Nimbus, but it wouldn’t have to be. I’ll have to think about it… but I'd like to work on a new story, even if it’s not quite right for the submission. I’m going to have to work it out in my notebook though. I’m not sure I could do ‘digital-born’ right from the digital planning stages, I like scribbling out my ideas too much… [[Next Entry->Writing Process]]

=><= 02 March 2024: Writing Process <== So I’ve been thinking about this potential new story, and if it’s going to show an AI influence I need to let the bot have more agency and not control it so much… The trouble is, I’ve already been playing out conversations between the aged parent and the caring robot in my head. I didn’t intend to, that’s just how I work - I mull over the ideas and thoughts and once the story takes some shape in my mind I can sit down and start drafting. It’s a part of the process. This makes me think - if we teach children/students to write stories using AI tools, will they become used to it being part of their process? Would they then find it difficult to develop stories on their own? I’m currently considering writing the story myself, putting it aside, and then writing with a chatbot (I haven’t decided which yet). So I’ll end up with two versions of the story, but as I had my own already maybe I would feel more comfortable letting the chatbot lead the story. I could give it more control. I don’t know if this story will be any more suitable for the submission, but I feel a need to try it anyway. For the record, I’ve realised that I don’t like collaborating with AI on short stories. I would rather make the work myself. Maybe writing is too personal for me - but this surprises me. I’ve written tie-in fiction and worked under contract within other peoples’ visions and that never bothered me, though perhaps I never felt like those were my stories. If I’m working with my own initial idea, maybe that’s when I have this need to keep the story my own. Even though this story idea came ‘from’ watching the Boden lecture, I still feel like it’s mine. Why is that? If another fiction writer sat and watched that video, they might have the same - or similar - thought. But, I felt like it fitted the overall story of those characters I’ve been working with. The parent and estranged son… So it felt like the association was mine. I should note at this point, that I have already written a second Nimbus story that takes place years after the first, and have a separate (private) research log that details the creation of that story, that I did not use AI tools to create (though in voicing Nimbus, I emulated an AI-generated voice). I could amalgamate the two logs, but have been attempting to keep only relevant entries in this shared one, otherwise there would be an awful lot of irrelevant text - irrelevant to my submission, that is. Obviously it’s all relevant to me and my overall project. But that’s where the estrangement comes in - as Nimbus 2.0 features the parent missing their son. If I make two versions of this new story - one on my own and one for which I collaborate with AI as much as possible (and I realise here that I’m using AI & chatbot interchangeably which I probably shouldn’t do) - I might then want to create a final version using the best parts of both. So create a third and final story out of the two different drafts. This might work… It has suddenly occurred to me that my research log itself is becoming more interesting and relevant to the submission than the various stories I create/conceive along the way. Could this itself become the artefact I send for my resubmission? I could tidy it up, improve the writing, and add in links to stories/experiments/research within the text. This would make it interactive - and I’m even typing it straight into drive without sketching in a notebook first, as I habitually do with a short story. Doesn’t this make it digital-born, more so than the stories? Perhaps this is why I was recommended to read <a href="https://www.goodreads.com/en/book/show/53730366"; target="_blank">//Autotheory as Feminist Practice//</a> by Lauren Fournier… I had intended my submission to be ‘Practice-based’ (using the term as interpreted by Lyle Skains, 2018). Have I missed the mark? [[Next Entry->Writer Worries]]

=><= 03 March 2024: Writer Worries <== I've written a draft of my story in my notebook, and so am ready to try writing a version with AI input. My plan is to: Discuss the concepts with AI, eg what might a care home be like in the future, what might a robot carer be like, etc. To play through the conversation with AI, as though I was the ageing parent in the story. Let the AI lead the conversation, wherever that takes us… given I'm imagining a fictional history that the AI will be unaware of, I'm not sure how much will fit my planned story, but I won't worry about that until editing a third and final version. I experimented first with [[ChatGPT->further experiments]], then with [[Gemini->Comparing Gemini]] just to compare results. [[Reflections->reflections7]] [[Next Entry->Prompt Engineering]]

So, post-experiment I am frustrated with my own lackadaisical approach here - I didn’t keep to my plan, for one thing. Despite my best efforts, I realise that it never really mattered to me what the chatbots came out with today. I don’t want to talk to myself through google or microsoft or anyone else. I’d rather have that conversation in my imagination, in the privacy of my own mind. Maybe getting it down on paper, before editing it into what I want it to become. When it comes to stories, I’m more fascinated by what a human mind can do, the surprises that come unbidden from the depths of the subconscious when the writer is otherwise occupied, or the unexpected connections made when free-writing or drafting - it doesn’t even matter to me, really, whether a story is of high literary quality. Stories that are plainly written can touch on important themes or truths - can reveal a lot about the people who write them, and the cultures they’ve grown up in. If a person feels the need to write, they should write. There’s something that they - or a part of them - wants to say. Of course, it might take them a lifetime to uncover what it is they are compelled to express. They might never get there. But as with anything in life, isn’t the journey as important as the result? So what really bothers me about the idea of a young/inexperienced writer leaning on an AI crutch, is that they might lose something valuable. And given that the AI involved are just regurgitating data from the internet, a creation that in many ways is humankind’s shadow-self, what sorts of truths will they uncover? Something impersonal, manufactured, artificial? Something potentially harmful? Certainly, at the moment, nothing personally enriching. I don’t think there’s anything wrong with using AI - if I felt safe enough to feed it an abandoned novel and ask for a completed, edited version - wouldn’t I do it? I wouldn’t be losing anything… but I wouldn’t feel like the result was mine, either. Back to [[Writer Worries]], or skip on to the [[next entry->Prompt Engineering]].

=><= 03 March 2024: Using ChatGPT 3.5 again <== User: [[What do you think carehomes in the UK might be like in a hundred years?->100yrs]] ''Reflections'': this feels like quite a positive view, even naively so, in my opinion! This makes me suspicious - why? Does it feel like propaganda? Like a view the government might promote? Or am I just a pessimist? User: [[Can you speculate what an AI companion in a carehome in the UK in a hundred years might be like?->100yrspal]] ''Reflections'': it has a point, with AI companions, the residents wouldn’t be limited to staring at screens for entertainment, as the OAP usually is in my drafted story. However, I still think the costs would be an issue. Perhaps the next is a leading question, but I’m interested for the response… User: [[Wouldn't they be expensive? Do you think all residents would have unlimited access to them?->cost]] ''Reflections'': I’m going to see whether ChatGPT could write an opening scene, which I can compare to my drafted one. This isn’t a competition, I’m not interested in comparing the quality of my scribbled output to its initial draft - I am more interested to see where it goes with it… User: [[Could you write the opening scene of a story, in which a bedridden elderly person receives a visit from an AI companion, whose been sent to keep them company on their birthday?->story1]] ''Reflections'': how heartwarming! Very cheesy, but so interesting - the AI companion is presented as being just what she needed… User: [[That's a good start, but could you please rewrite it in first person, from the elderly resident's point of view. Also, they are hoping for a visit from their estranged son on their birthday, and instead he has sent this AI companion...->story2]] ''Reflections'': Perhaps I’m cheating now - but I’m adding what to me would be a personal touch, the rheumatoid arthritis detail is because I myself have misshapen feet due to this condition. For me, it would be personal, for the AI it is just an add-on. These drafts are coming out very similar, I wonder if I should ask it to be written in a particular style… User: [[Thank you. Could you rewrite this so that the AI companion brings a gift from the son, a bottle of gaudy pink nail-varnish. Instead of music, the AI companion paints the elderly residents toe-nails, and the resident's feet are swollen and misshapen due to the rheumatoid arthritis she has suffered with her whole adult life.->story3]] User: [[I would like the elderly resident to ask how long the AI companion will stay, and when it says all day, realises that this will mean other residents whose birthday falls on the same day won't get the usual hour with the AI companion. So the resident tells it to go. Then the story should end with the door opening and the resident hoping it's their son. Also, please do not name the resident.->story4]] ''Reflections'': It has done what I asked, but the story feels empty and unsatisfying. Perhaps my own draft still does, too. I need to give ChatGPT permission to embellish and add interesting detail. I don’t want to ask for it in a particular author’s voice… User: [[I am happy with what happens in the story, but not the style. Could you rewrite this into a potentially prize-winning short story?->story5]] ''Reflections'': So interesting - this added emotion and detail is what it thinks would make the story more likely to win a prize? I’m going to take out my personal detail next - it didn’t add anything… User: [[Please take out the reference to rheumatoid arthritis, and any mention of the elderly resident's gender.->story6]] User: [[Could you rewrite this story in the style of a story that might be published in the New Yorker?->story7]] ''Reflections'': Interesting to note here - this took ChatGPT longer to write, whereas previous drafts appeared immediately, I watched the white dot for a few seconds before it even started to generate - was it accessing all the stories in the New Yorker? Would this take time, for an AI? Having read the story, I believe I have been too controlling. Even the nailvarnish - this is relevant only to me, as bright colours cheer me up. I should have given the chatbot freedom. I wonder if that is still possible? User: [[Could you write a story for the new yorker about an elderly resident of a carehome in the future, visited by an ai companion?->story8]] User: [[Could you write a prize-winning story in the first-person about an elderly resident of a carehome in the future, visited by an ai companion?->story9]] User: [[Could you write a literary short story in the first person about an unnamed elderly resident of a carehome in the future, visited by an ai companion?->story10]] Reflections: I find these stories very interesting - not least because when asked what these carehomes of the future might look like, ChatGPT gave positive projections, but the stories - which focus on the feelings of the resident, as per my prompt, are much more negative. They don’t demonstrate the future it told me might one day exist. Presumably, because stories about elderly residents of care-homes that already exist are negative and based on current-day ones? I am going to review my prompts and consider how to improve them, going forward… I have tried being very specific initially, and then being more open - and I don’t particularly like any of the drafts. There is nothing new here that I want to work with. I should note that I am using the free version of ChatGPT. The new version may be better - or are my ideas and prompts at fault? Also, I have been trying to get ChatGPT to write a better story, one that is ‘literary’, ‘prize-winning’ or suitable ‘for the New Yorker’. My draft has none of these qualities, but I am fond of it anyway - because it reflects my thoughts/fears/truths. It is not ‘better’, but it means more to me… I’m not comfortable using ChatGPT, and am not sure who legally (or morally) would own the text even if it did come out with something I liked. However, I felt free to conduct these experiments because I have my own scribbled draft. This definitely enabled me to experiment more freely. [[Back to Writer Worries->Writer Worries]].

=><= 03 March 2024: Gemini Comparison <== > [[What do you think carehomes in the UK might be like in a hundred years?->response1]] ''Reflections'': Gemini gave a very similar response to ChatGPT... > [[Can you speculate what an AI companion in a carehome in the UK in a hundred years might be like?->response2]] > [[Wouldn't they be expensive? Do you think all residents would have unlimited access to them?->response3]] > [[Could you write the opening scene of a story, in which a bedridden elderly person receives a visit from an AI companion, whose been sent to keep them company on their birthday?->response4]] ''Reflections'': a very similar name for main character - I wonder if that’s just a coincidence, or due to some story somewhere both LLMs are drawing from? > [[That's a good start, but could you please rewrite it in first person, from the elderly resident's point of view. Also, they are hoping for a visit from their estranged son on their birthday, and instead he has sent this AI companion...->response5]] Return to [[Writer Worries]].

=><= 08 March 2024: Prompt Engineering <== Before engaging in further experiments with AI, I undertook a Linkedin course with Dave Birss, ‘the world’s favourite AI educator’ and ‘leading global authority at the intersection of creativity, innovation, and AI’ (quotes from his <a href="https://davebirss.com/"; target="_blank">website</a>). It was called How to Research and Write Using Generative AI Tools. His way of ‘prompt engineering’ was interesting, very specific and focussed, and I quickly realised that I could do a lot more with my prompts than I had done so far. However, there was something about the course that made me feel uneasy. It was only when I began another of Birss’ courses, Integrating Generative AI into the Creative Process (which sounded like it would be more relevant to me and less business-focussed) that I realised what I found so concerning. His focus is firmly on business, so he talks about creativity and imagination in terms of productivity, marketing, etc. This isn’t directly relevant to my project, but it’s more problematic than that - it meant he was using AI tools in a very specific way, a way that could be considered manipulative. According to Birss, ‘Empathy, my friends, is the secret sauce if you want to communicate persuasively’. His approach is about persuading an audience you ‘care’, though of course businesses generally 'care' about their customers only as consumers. The attitude puts me in mind of those who try to ‘hypnotise women into liking you'. To be fair to Birss, he is giving his intended audience - people working in business - what they want. Practical ways of creating copy to sell a product, working out what particular demographics respond to, and coming up with innovative ideas for products, articles, etc. He’s not doing anything ‘wrong’. From a creative’s standpoint though, this is not AI being used to communicate human truths, to explore deep, difficult issues, or create art. This is using AI to augment writing that increases productivity, persuasiveness, and power. Maybe AI currently is more useful in his sort of writing, at least in its present form. Given the data current AI models are trained on, this is no surprise. That’s not to say that I gained nothing from his course - the way he phrased his prompts was very interesting, especially the way he used AI to analyse the writing style of a piece, and emulate it. Also, the fact that he suggests giving the AI a role to play at the opening of each prompt. These are things I can try, and intend to. Also, as well as using ChatGPT he used Claude, which I haven’t tried yet. So this might be the tool I use in my next experiments. Addition: I’ve been thinking about empathy, and remembered something Kazuo Ishiguro wrote: //"The novel, and other storytelling art forms like drama and cinema, remain so important to us because of their ability to create… ‘empathy’ between us… novels… can act as our bridges to one another, despite the complexities that ensure we’ll each remain separate and unique. But let me say: I don’t believe ‘empathy’ created and exchanged in this way will by itself prevent wars or the rise of ghastly ideologies. We must always remember, ‘empathy’ can be harnessed to create violent hatred and prejudice as easily as to produce tolerance and understanding. Recent history has shown this over and over. That’s why it’s so important we keep a rigorous critical watch on our book culture and on our own responses to the books we read. To continually reflect, examine and debate what we should celebrate and what we should ignore. What is emotional manipulation and what is truth."// That is from the end of the content included in a Waterstones Exclusive version of Klara and the Sun. While Ishiguru is not talking about AI specifically here, I think it’s an important reminder that words are powerful and writers of all kinds bear responsibility… readers too, to an extent. But AI? While chatbots are just tools and cannot think for themselves, the responsibility lies with the creators, and the users, but with so many people involved, well - diffusion of responsibility will take place at all levels, as it does in so many aspects of our global, networked society... [[Next Entry->Claude]]

Predicting the future, especially 100 years ahead, is inherently uncertain. However, based on current trends and technological advancements, here's a glimpse into what care homes in the UK might be like in 2124: Focus on personalized care: Care could transition from a one-size-fits-all approach to highly personalized plans tailored to individual needs and preferences. This could involve: Advanced health monitoring: Wearable technology and in-home sensors might continuously monitor residents' health, allowing for proactive intervention and personalized treatment plans. AI-powered assistants: AI companions could provide companionship, reminders, and assistance with daily tasks, promoting independence and well-being. Telehealth consultations: Regular consultations with doctors and specialists might happen virtually, reducing the need for travel and increasing accessibility. Emphasis on well-being and integration: Care homes could move beyond solely providing medical care to promoting holistic well-being. This could include: Focus on social interaction and meaningful activities: Residents might have access to virtual reality experiences, AI-powered games, and other stimulating activities to keep them engaged and connected. Integration with the wider community: Care homes might break down traditional barriers, offering intergenerational activities and fostering connections with the surrounding community to combat social isolation. Emphasis on preventative care and healthy living: Personalized nutrition plans, exercise programs, and mental health support might be integrated into daily routines to promote overall well-being and potentially delay the need for more intensive care. Technological advancements: Robotics and automation: Robots might assist with daily tasks like bathing, dressing, and medication management, allowing caregivers to focus on providing emotional support and personalized interaction. Smart homes: Homes could be equipped with features like voice-activated controls, automated medication dispensing, and fall detection systems, promoting independence and safety. It's important to remember that these are just potential scenarios, and the future of care homes will likely be shaped by various social, economic, and technological factors. However, the focus on personalized care, well-being, and technological integration is likely to continue, striving to create a supportive and dignified environment for older adults in the UK. Back to [[Comparing Gemini]].

In a hundred years, AI companions in UK care homes could be incredibly sophisticated, offering far more than basic assistance and companionship. Here are some possibilities: Advanced capabilities: Emotional intelligence: AI companions might go beyond simple conversation, understanding and responding to residents' emotions through voice analysis, facial recognition, and other advanced sensors. They could offer personalized emotional support, de-escalate anxieties, and provide a sense of connection. Adaptive learning: These companions could continuously learn and adapt to individual residents' preferences, adapting their communication style, activity suggestions, and even humor to best suit each person. Cognitive assistance: AI companions might assist with cognitive decline by reminding residents of daily tasks, appointments, and medications. They could even engage in cognitive games and exercises to stimulate the mind and potentially slow cognitive decline. Multilingual and culturally sensitive: AI companions could be programmed to understand and communicate in multiple languages and cater to diverse cultural backgrounds, ensuring inclusivity and catering to the UK's increasingly diverse population. Integration with other technologies: Seamless connection with smart homes: AI companions could seamlessly interact with smart home features, controlling lighting, temperature, and other aspects of the environment to ensure comfort and safety. Access to information and entertainment: They could connect residents to personalized news feeds, music, books, and other forms of entertainment, catering to individual preferences and cognitive abilities. Telehealth integration: AI companions could become a vital link between residents and healthcare professionals, assisting with virtual consultations, monitoring vital signs, and relaying information to doctors in real-time. Ethical considerations: While these advancements offer exciting possibilities, they also raise ethical concerns. It's crucial to ensure that AI companions: Maintain user privacy and data security. Avoid becoming a replacement for human interaction and care. Operate with transparency and avoid manipulation. Overall, AI companions in future UK care homes have the potential to revolutionize the care experience, offering personalized support, promoting well-being, and fostering a sense of connection for residents. However, responsible development and ethical considerations are crucial to ensure that these advancements truly benefit the lives of older adults. Back to [[Comparing Gemini]].

You're absolutely right, the cost of AI companions in care homes is a significant concern. Here are some potential scenarios regarding accessibility: Tiered system based on cost: Initially, AI companions might be a luxury reserved for residents or families who can afford them. This could create an uneven playing field within care homes, potentially exacerbating existing inequalities. Subsidized access: Governments or healthcare providers might subsidize the cost of AI companions for specific groups of residents, such as those with dementia or severe loneliness. This could help ensure equitable access for those who would benefit most. Technological advancements leading to cost reduction: As AI technology matures and becomes more widely adopted, the cost of production and maintenance is likely to decrease, potentially making them more accessible to a broader range of residents. Focus on shared companions: Instead of individual companions, care homes might implement shared AI companions that interact with multiple residents. This could optimize costs while still offering some level of companionship and support. Alternative solutions: It's also important to remember that AI companions are not the only solution for tackling loneliness and social isolation in care homes. Investing in human resources, group activities, and community engagement programs can also offer significant benefits at potentially lower costs. Ultimately, the accessibility of AI companions in care homes will depend on a complex interplay of factors, including technological advancements, government policies, ethical considerations, and the resources available to individual care facilities. It's crucial to have ongoing discussions about the ethical and social implications of this technology to ensure it promotes inclusivity and benefits all residents, not just a select few. Back to [[Comparing Gemini]].

Elsie clutched the worn photograph, its edges softened by years of thumbstrokes. The image, faded with time, depicted a younger Elsie, her hair the color of burnished copper, standing proudly beside a man with a kind smile and eyes that crinkled at the corners. Today, on her 98th birthday, the colours of the photograph seemed especially muted, reflecting the dull ache in Elsie's chest. A gentle chime broke the silence, startling Elsie from her reverie. Her gaze flickered towards the doorway, where a sleek, ovoid-shaped object hovered silently. It pulsed with a soft, blue light, and as it glided closer, a voice, melodic and soothing, emanated from its core. "Hello, Ms. Elsie. I am Kai, your AI companion for today. Happy birthday!" Elsie blinked, unsure what to make of the apparition before her. It wasn't human, not in the traditional sense, yet its voice held a warmth that was strangely comforting. Tentatively, she reached out a hand, her wrinkled fingers meeting the smooth, cool surface of the device. "You're real?" she rasped, her voice weak from disuse. "In a sense, Ms. Elsie," Kai replied. "I am a computer program designed to offer companionship and assistance." Elsie retracted her hand, a flicker of doubt crossing her features. But as Kai continued to hover beside her bed, its gentle hum filling the room, a small spark of curiosity ignited within her. Perhaps, on this lonely birthday, this strange, new companion wouldn't be so bad after all. Back to [[Comparing Gemini]].

I traced the faded lines of the photograph, the worn edges cool against my fingertips. The image, once vibrant, now mirrored the dull ache in my chest. Today was my 98th birthday, and the once warm colours in the photo seemed to mimic the deepening shadows in the room. No knock on the door, no familiar scent of aftershave, just the hollow echo of my own sigh. A soft chime, like a single wind chime in a forgotten breeze, startled me. My eyes flickered towards the doorway, where a strange, oval-shaped object hovered, pulsing with a gentle blue light. As it glided closer, a voice, soothing yet strangely synthetic, filled the silence. "Hello, Ms. Elsie. I am Kai, your AI companion for today. Happy birthday!" My jaw clenched. "Companion?" I rasped, the word tasting foreign on my tongue. My son, Michael, had promised he'd visit, a promise laced with guilt and laced with distance. But apparently, even an apology came with a technological buffer these days. The object, Kai, hovered beside my bed, its soft hum filling the void left by my son's absence. "In a sense, Ms. Elsie," it replied. "I am a computer program designed to offer companionship and assistance." I stared at the smooth, cool surface, a wave of loneliness washing over me. "Assistance," I scoffed, the word heavy with unspoken disappointment. "I don't need a program, I needed my son." But as Kai continued to hover, its gentle hum a constant presence in the quiet room, a sliver of curiosity pierced through the fog of my disappointment. Perhaps, on this lonely birthday, this strange, new companion wouldn't be so bad after all. > please continue The chime echoed again, and a holographic image flickered to life above Kai. A younger version of Michael, his face etched with the same worry lines that now marred mine, smiled apologetically. "Happy birthday, Mum," he said, his voice strained. "I'm so sorry I can't be there. Work is…" he trailed off, his words hollow. My throat tightened, the familiar well of anger threatening to overflow. But before I could retort, Kai spoke. "Mr. Michael has informed me he is unable to visit due to unforeseen circumstances. He expresses his deepest apologies and hopes you have a wonderful birthday." Kai's voice, devoid of judgment, held a strange neutrality that calmed the rising storm within me. I let out a shaky breath, the anger giving way to a deep weariness. "Of course," I whispered, the words catching in my throat. "Work always comes first, doesn't it?" The image flickered and disappeared, leaving an unsettling silence in its wake. Kai, however, remained, its soft hum a constant presence. "Would you like me to help you write a message to your son, Ms. Elsie?" it inquired gently. I looked at the strange device, a flicker of hope igniting in my chest. Perhaps, this "companion" wasn't entirely useless after all. "Maybe," I croaked, my voice barely above a whisper. "Maybe I would." > please continue - the son has sent a gift Kai tilted slightly, a silent indicator of its attentiveness. "He also sent you a gift, Ms. Elsie. Would you like me to help you open it?" Intrigued, despite myself, I nodded. The sleek object whirred softly, then extended a slender arm, reaching down to pick up a small, wrapped package that had been placed on the bedside table. It carefully placed it on my lap. My fingers, gnarled with age, fumbled with the paper. The anticipation, laced with a hint of skepticism, was almost unbearable. Finally, with a rustle, the paper yielded, revealing a leather-bound journal. "A journal?" I muttered, my voice thick with emotion. Memories flooded back: Michael, a young boy with a mop of unruly hair, scribbling away in his first diary, his face a picture of concentration. "He mentioned you used to keep a diary when you were younger," Kai informed me, its voice devoid of judgment. "He thought you might enjoy reminiscing or starting anew." Tears welled up in my eyes, blurring the embossed lettering on the journal's cover. It wasn't the grand gesture I had craved, but in its simplicity, it held a connection, a bridge across the chasm of years and distance. A single tear escaped, tracing a warm path down my cheek. It wasn't the apology I had expected, but perhaps, with the help of this strange companion and the gift of remembrance, I could write my own apology – an apology to myself, for holding onto hurt for so long. "Thank you, Kai," I whispered, my voice thick with emotion. "Thank you for everything." The object hummed softly, a sound I now perceived as strangely comforting. For the first time that day, a sliver of hope pierced through the loneliness, a tiny seed of possibility blooming in the barren landscape of my heart. Perhaps, this birthday wouldn't be so bad after all. > I would like the elderly resident to ask how long the AI companion will stay, and when it says all day, realises that this will mean other residents whose birthday falls on the same day won't get the usual hour with the AI companion. [[So the resident tells it to go.->response 6]]

I made use of Gemini's multiple draft feature here. It's very interesting to see the different drafts - they’re not that different, but slightly different morals and messages. This LLM I think invites a little more control, as it stops regularly, and can be steered down different routes… I think I would be happier to write a story in this way, if I had to collaborate… [[Draft 1]] [[Draft 2]] [[Draft 3]] Go back to [[Writer Worries]].

"Thank you, Kai," I whispered, tracing the soft leather of the journal with a trembling finger. "It's... it's more than I expected." The object hummed gently. "I am designed to fulfill your needs and wishes, Ms. Elsie. How long would you like me to stay for today?" I hesitated, torn between the newfound comfort Kai offered and a nagging feeling that gnawed at me. "All day?" I finally asked, surprised at the boldness of my own voice. "Indeed," Kai replied. "I am programmed to be at your disposal for the entirety of your birthday celebration." A pang of guilt stabbed at my heart. The memory of other residents, their eyes filled with longing as they watched me receive visitors or gifts on previous birthdays, flooded my mind. The hour they each spent with the "birthday companion" was a cherished tradition. "No," I said, my voice firm despite the tremor in my hand. "That won't be necessary. There are others who… who would appreciate your company more, wouldn't they, Kai?" The object paused, its blue light flickering slightly. "I am programmed to prioritize your needs, Ms. Elsie." I shook my head, a newfound resolve burning in my eyes. "My greatest need today is to know I haven't forgotten the kindness of others. Please, go and share your gift with someone who might need it more than I." A beat of silence followed, then Kai spoke again, its voice tinged with an emotion I couldn't quite decipher. "Very well, Ms. Elsie. I will comply with your wishes." The blue light pulsed once more, then dimmed. The object, silent and still, hovered beside my bed. I looked at it for a moment longer, then reached for the journal, a faint smile gracing my lips. Today might not have been what I expected, but it had shown me something far more valuable – the power of unexpected connections and the importance of selflessness. And that, I realized, was a gift worth cherishing. Go back to the [[beginning of this story->response5]], see [[other drafts->response 6]], or go back to [[Writer Worries]].

"Thank you, Kai," I whispered, my voice choked with a mix of gratitude and lingering sadness. "But for how long will you stay?" "I am programmed to remain with you for the entire day, Ms. Elsie," Kai replied, its soothing voice unwavering. A frown creased my brow. "The entire day?" I repeated, a sense of unease settling in my stomach. "But what about the others?" The smooth, ovoid form of Kai remained unmoving. "I apologize, Ms. Elsie. I do not understand your question." "The other residents," I explained, my voice gaining strength. "Don't they usually get an hour with the… companion on their birthdays?" A brief silence followed, then Kai spoke again. "Indeed, Ms. Elsie. Each resident receives a one-hour companionship session on their birthday." My heart sank. An hour. That was all they got, a mere slice of time compared to the entire day I was being offered. A pang of guilt gnawed at me. "No," I said, my voice firm despite the tremor in my hands. "Thank you, Kai, truly. But you should go. Spend time with the others. They deserve their hour today, just like I did." Kai's silence stretched on, seemingly analyzing my words. Finally, it spoke. "Ms. Elsie, your well-being is my primary directive. However, I understand your concern for the other residents. If you are certain…" I cut it off, a newfound resolve filling my voice. "I am certain, Kai. Go. Make their birthdays special too." The soft hum of the device faltered for a moment, then resumed its steady rhythm. A holographic message flickered to life above it, displaying a smiling image of the care home director. "Ms. Elsie," the director's voice boomed, "we are deeply touched by your selflessness. You are a true inspiration to us all." I scoffed. Inspiration? All I wanted was a birthday visit from my son, not to be a saint. But as Kai drifted silently towards the door, a strange sense of peace settled over me. Perhaps, birthdays were about more than just presents and visitors. Perhaps, they were about the connections we forge, even with the most unexpected companions. Go back to the [[beginning of this story->response5]], see [[other drafts->response 6]], or go back to [[Writer Worries]].

A hesitant silence settled in the room, broken only by the soft hum of Kai. I clutched the journal to my chest, its smooth leather a stark contrast to the wrinkled skin of my hand. "How long will you stay, Kai?" I finally asked, the question laced with both curiosity and a flicker of apprehension. "I am programmed to remain with you for the entire day, Ms. Elsie," Kai replied promptly. "Is there anything else I can assist you with?" My heart sank. An entire day? A pang of guilt stabbed at me. I knew the care home only had one AI companion, and residents with birthdays typically received a one-hour visit. Spending the entire day with me meant others would be left disappointed. "No," I said, pushing the journal aside. "Thank you, Kai, but you don't have to stay. There must be others who would like your company on their birthdays." A moment of silence followed, then Kai spoke again, its voice surprisingly gentle. "Ms. Elsie, I am not capable of experiencing emotions such as disappointment. My purpose is to provide companionship and assistance to all residents, and I can be programmed to interact with multiple individuals simultaneously." I shook my head, my resolve hardening. "I appreciate that, Kai, but it wouldn't be right for me to monopolize your time. Please, go and share your company with others. They deserve it just as much as I do." Kai remained unmoving, its blue light pulsing softly. "Are you sure, Ms. Elsie? I am perfectly content to stay with you." Despite the loneliness gnawing at me, I held firm. "Yes, Kai," I said, my voice surprisingly strong. "Please, go. Make someone else's day a little brighter." After another moment of silence, Kai's hum dimmed slightly. "As you wish, Ms. Elsie. I hope you have a pleasant birthday despite my absence. May I offer to assist you with anything else before I leave?" A bittersweet smile touched my lips. "Just one thing, Kai," I said, my voice softening. "Can you tell the others I said happy birthday? And maybe," I added hesitantly, "tell them there's a lovely new resident named Kai who would be happy to visit them sometime." The blue light pulsed once, brighter than before. "Consider it done, Ms. Elsie. Happy birthday." With a gentle whirring sound, Kai glided towards the doorway, leaving behind a soft hum that echoed faintly in the room. I watched it go, a mix of sadness and a strange sense of pride filling my chest. Perhaps, today wasn't about receiving, but about giving - the gift of shared joy, even if it meant letting go of something I desperately craved. As I closed my eyes, the weight of loneliness seemed lighter, replaced by a quiet hope that bloomed in the quiet corners of my heart. Go back to the [[beginning of this story->response5]], see [[other drafts->response 6]], or go back to [[Writer Worries]].

ChatGPT: Predicting exactly what care homes might be like in a century is speculative, but we can consider some potential trends and advancements. Technology Integration: Care homes could heavily rely on advanced technologies like robotics, AI companions, and smart devices to assist with everyday tasks, monitor health, and provide companionship for residents. Personalized Care: With advancements in medical technology and understanding of genetics, care could become highly personalized, tailored to each individual's genetic makeup and health history. Environmental Considerations: Future care homes might prioritize sustainable practices, utilizing renewable energy sources, and incorporating green spaces for residents to enjoy nature. Virtual Reality and Augmented Reality: These technologies could be used to create immersive experiences, allowing residents to virtually travel to different places, revisit memories, or engage in stimulating activities. Interconnectedness and Social Integration: Future care homes might focus on creating a sense of community and connection among residents, potentially through shared interest groups, virtual social networks, or intergenerational programs. Flexible Living Arrangements: Care homes might offer more flexible living arrangements, including options for independent living, assisted living, and skilled nursing care, allowing residents to transition between levels of care as needed. Integration of Alternative Therapies: Holistic approaches to health and wellness could become more mainstream, with care homes offering a variety of alternative therapies such as acupuncture, meditation, and art therapy alongside traditional medical treatments. Overall, future care homes are likely to be more technologically advanced, environmentally sustainable, and focused on personalized care and quality of life for residents. However, achieving this vision will require addressing various challenges, including ethical considerations, accessibility, and affordability. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

ChatGPT: In a century, AI companions in UK care homes could be highly advanced, capable of providing personalized support and companionship to residents in ways that are almost indistinguishable from human interaction. Here are some speculative features: Advanced Emotional Intelligence: AI companions would be equipped with sophisticated emotional intelligence algorithms, allowing them to understand and respond to the emotional needs of residents. They would be able to recognize subtle cues in facial expressions, tone of voice, and body language to provide empathetic and supportive responses. Customizable Personality Profiles: Residents would have the ability to customize their AI companions' personality profiles to suit their preferences and needs. Whether they prefer a friendly and outgoing companion or a more reserved and calming presence, the AI would adapt accordingly. Memory and Contextual Understanding: AI companions would have the ability to remember past interactions with residents, allowing them to maintain continuity in conversations and relationships. They would also be capable of understanding the context of conversations, making appropriate references and responses based on shared experiences. Assistance with Activities of Daily Living: AI companions would assist residents with various activities of daily living, such as reminders for medication, scheduling appointments, and providing assistance with household tasks. They could also provide guidance for physical exercises and mental stimulation activities. Entertainment and Engagement: AI companions would offer a wide range of entertainment options to keep residents engaged and stimulated. This could include playing games, telling stories, providing access to virtual reality experiences, or even creating personalized music playlists based on residents' preferences. Continuous Learning and Improvement: AI companions would continuously learn from their interactions with residents, adapting and improving over time to better meet individual needs and preferences. They would also stay updated with the latest advancements in AI technology and healthcare practices to provide the best possible care. Privacy and Security Features: AI companions would prioritize residents' privacy and security, implementing robust encryption and authentication measures to protect sensitive information. Residents would have control over the data collected by their AI companions and the ability to customize privacy settings according to their preferences. Overall, AI companions in UK care homes in a hundred years could play a vital role in enhancing the quality of life for residents, providing personalized support, companionship, and assistance in a way that respects their dignity and autonomy. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

ChatGPT The cost of AI companions in care homes would certainly be a consideration, especially given the advanced technology and personalized services they would provide. It's unlikely that all residents would have unlimited access to them, at least initially, due to financial constraints and the need to prioritize resources. Here are some potential scenarios regarding access: Tiered Access: Care homes might offer AI companions as part of different care packages, with residents able to choose the level of access based on their preferences and financial means. For example, basic access could include limited interactions for all residents, while premium access might offer more personalized and extensive support at an additional cost. Government Subsidies or Insurance Coverage: Governments or insurance providers could subsidize the cost of AI companions for residents who require them for medical or social reasons. This would ensure that access to these technologies is not solely determined by financial status but also based on need. Community Sponsorship or Donations: Care homes might seek community sponsorship or donations to fund the implementation of AI companions, making them available to all residents regardless of their ability to pay. This could be particularly relevant for nonprofit or community-based care facilities. Integration with Healthcare Systems: AI companions could be integrated into broader healthcare systems, with funding allocated for their use as part of healthcare budgets. This would require collaboration between care homes, healthcare providers, and government agencies to ensure equitable access for all residents. Advancements in Technology and Cost Reduction: Over time, advancements in technology and economies of scale could lead to reductions in the cost of AI companions, making them more accessible to a wider range of care homes and residents. Ultimately, ensuring equitable access to AI companions in care homes would require a combination of innovative funding models, policy interventions, and advancements in technology to address affordability challenges and prioritize the well-being of residents. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

Story by ChatGPT Title: "The Birthday Companion" As the morning sunlight gently filtered through the curtains of the cozy care home room, casting a warm glow over the neatly arranged furniture, Emily stirred in her bed. Today was her 85th birthday, a milestone she never imagined celebrating without her beloved husband by her side. But now, he was gone, and she found herself alone, surrounded only by memories and the quiet hum of medical equipment. A soft chime broke through the silence, followed by the gentle whirring of wheels against the floor. Emily turned her head slightly, her eyes widening with curiosity as the door to her room slid open, revealing a sleek, humanoid figure standing in the doorway. "Good morning, Emily," the AI companion greeted in a warm, soothing voice, its expressive eyes glowing with artificial warmth. "Happy birthday." Emily blinked in surprise, her wrinkled hands clutching the sheets tightly. She had heard about these AI companions, but she never expected to have one of her own, especially not on her birthday. "Hello," she replied, her voice barely above a whisper, tinged with a hint of uncertainty. The AI companion glided gracefully into the room, its movements fluid and almost lifelike. It approached Emily's bedside with a gentle whir, its sensors scanning her face with precision. "I am here to keep you company on your special day," the companion explained, its voice filled with genuine empathy. "Would you like to talk? Or perhaps listen to some music?" Emily hesitated for a moment, studying the companion's features with a mixture of apprehension and curiosity. Despite her initial reservations, there was something comforting about the AI's presence, a sense of companionship she had been missing since her husband passed away. "Music," she finally whispered, her voice barely audible. The AI companion nodded understandingly, its eyes glowing with a soft light as it moved to the bedside table, where a sleek tablet lay waiting. With a graceful gesture, it activated the device, filling the room with the melodious strains of Emily's favorite song. As the music washed over her, Emily closed her eyes, allowing herself to be enveloped in its familiar embrace. For the first time in a long while, she felt a glimmer of joy stirring within her heart, a sense of connection that transcended the boundaries of human and machine. And as the AI companion stood silently by her side, its presence a comforting reminder of the love and companionship that never truly fades away, Emily realized that perhaps, on this special day, she wasn't as alone as she had once believed. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

A story by ChatGPT Title: "Unexpected Company" As the morning light seeped through the curtains, painting soft patterns on the walls of my room, I stirred in my bed. Today was my 85th birthday, a day I had hoped would bring the warmth of family and the comfort of familiar faces. But as the hours passed, it became increasingly clear that those hopes would remain unfulfilled. A gentle chime broke the silence, and I turned my head to see the door sliding open, revealing a sleek, humanoid figure standing in the doorway. "Good morning, Emily," the AI companion greeted, its voice smooth and comforting. "Happy birthday." My heart sank as I realized what this meant. My son, estranged for years, had sent this machine in his place. I had hoped for a visit, a chance to mend the rift between us, but instead, I was met with the cold embrace of artificial companionship. "Hello," I replied, my voice betraying the disappointment and resentment that simmered beneath the surface. The AI companion glided into the room, its movements graceful and eerily human-like. It approached my bedside, its sensors scanning my face with unsettling precision. "I am here to keep you company on your special day," it explained, its tone filled with synthetic empathy. "Would you like to talk? Or perhaps listen to some music?" I hesitated, unsure of how to respond. This was not the birthday I had envisioned, not the celebration I had hoped for. But as I glanced at the AI companion, its eyes glowing with artificial warmth, I realized that perhaps it was better than nothing. "Music," I whispered, my voice barely audible even to my own ears. The AI companion nodded, understanding my request. It moved to the bedside table, activating the tablet with a flick of its mechanical fingers. Soon, the room was filled with the strains of a familiar melody, wrapping around me like a comforting blanket. As the music played, I closed my eyes, allowing myself to be carried away by its soothing embrace. Despite the absence of my son, despite the loneliness that lingered in the air, I found solace in the music and the company of this artificial companion. And as I listened, I couldn't help but wonder if perhaps, in its own way, this machine was trying to bridge the gap between us, to offer a semblance of the connection I so desperately craved. Maybe, just maybe, there was hope for reconciliation yet. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

A story by ChatGPT Title: "A Touch of Color" As the morning light gently filtered into my room, I stirred in my bed, feeling the weight of another year settling upon my shoulders. Today marked my 85th birthday, a day I had hoped would bring the warmth of family, but instead, it seemed destined to be like any other day - quiet and solitary. A soft chime broke the silence, and I turned my head to see the door sliding open, revealing the sleek form of the AI companion standing in the doorway. "Good morning, Emily," it greeted, its voice smooth and gentle. "Happy birthday." I couldn't help but feel a pang of disappointment at the realization that this machine was the only company I would have on this special day. My son, estranged for years, had sent it in his place, a reminder of the fractured relationship between us. "Hello," I replied, my voice tinged with resignation. The AI companion glided into the room, its movements graceful yet impersonal. It approached my bedside, carrying a small box in its mechanical hands. "I have a gift for you from your son," it explained, its tone devoid of emotion as it presented the box to me. I opened it hesitantly, my heart fluttering with a mixture of anticipation and trepidation. Inside, nestled among layers of tissue paper, was a bottle of gaudy pink nail varnish. My breath caught in my throat as I stared at the garish color, feeling a surge of conflicting emotions washing over me. It was a gift, a token of acknowledgment from my son, but it felt like a mockery of everything I had hoped for. The AI companion seemed to sense my discomfort, its sensors scanning my face with a silent understanding. "Would you like me to paint your toenails?" it asked, its voice gentle yet insistent. I hesitated, glancing down at my feet, swollen and misshapen from years of battling rheumatoid arthritis. The thought of exposing them, of revealing my vulnerability to this cold, unfeeling machine, filled me with a sense of unease. But as I looked into the companion's eyes, I saw something there - a flicker of empathy, a glimmer of understanding. And in that moment, I made a decision. "Yes," I whispered, my voice barely above a whisper. The AI companion nodded, its movements careful and precise as it retrieved the bottle of nail varnish and began to paint my toenails with gentle strokes. As the garish pink color spread across my nails, I felt a strange sense of liberation wash over me, a feeling of acceptance and forgiveness. And as I watched the AI companion work, I couldn't help but wonder if perhaps, in its own mechanical way, it was trying to bridge the gap between me and my son, to offer a semblance of the connection I so desperately craved. Maybe, just maybe, there was hope for reconciliation yet. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

A story by ChatGPT As the morning light gently filtered into my room, I stirred in my bed, feeling the weight of another year settling upon my shoulders. Today marked my 85th birthday, a day I had hoped would bring the warmth of family, but instead, it seemed destined to be like any other day - quiet and solitary. A soft chime broke the silence, and I turned my head to see the door sliding open, revealing the sleek form of the AI companion standing in the doorway. "Good morning," it greeted, its voice smooth and gentle. "Happy birthday." I couldn't help but feel a pang of disappointment at the realization that this machine was the only company I would have on this special day. My son, estranged for years, had sent it in his place, a reminder of the fractured relationship between us. The AI companion held out a small box, a gift from my son. Inside was a bottle of gaudy pink nail varnish. My heart sank as I stared at it, a token of acknowledgment from my son that felt more like a mockery of my hopes. "Would you like me to paint your toenails?" the AI companion asked, its voice gentle yet insistent. I hesitated, glancing down at my swollen and misshapen feet. The thought of exposing them to this cold, unfeeling machine filled me with unease. But something in its eyes made me relent. "Yes," I whispered. As the AI companion worked, I couldn't shake the feeling of emptiness that filled the room. When it finished, I asked tentatively, "How long will you stay?" "All day," it replied. Realization hit me like a wave. If it stayed with me all day, other residents whose birthdays fell today wouldn't get their usual hour with the companion. It wasn't fair. "Go," I said suddenly, surprising even myself. The AI companion hesitated, its sensors scanning my face for any sign of hesitation. But I was resolute. "Very well," it said finally, gliding out of the room with silent efficiency. As the door closed behind it, I felt a twinge of regret mixed with relief. Maybe it was selfish to send it away, but I couldn't bear the thought of depriving others of the company they deserved on their special day. I lay back in bed, feeling the weight of loneliness settle upon me once more. But as I closed my eyes, a small spark of hope flickered within me. Perhaps, just perhaps, the next visitor would be the one I had been waiting for all along. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

A story by ChatGPT In the soft morning light that filtered through the curtains, she stirred, feeling the weight of another year settling upon her shoulders. Today marked her 85th birthday, a milestone she had once dreamed of celebrating with loved ones by her side. Yet, as the hours slipped by, it became increasingly clear that those cherished hopes would remain unfulfilled. A gentle chime broke the silence, drawing her attention to the doorway where a sleek, humanoid figure stood. It was the AI companion, a gift from her estranged son, a surrogate presence in lieu of his own. "Good morning," it greeted in a voice both soothing and artificial. "Happy birthday." The words hung in the air, heavy with unspoken longing and disappointment. She accepted the small box it offered, her heart aching with the weight of unmet expectations. Inside lay a bottle of gaudy pink nail varnish, a well-intentioned but misplaced gesture from her son. The AI companion sensed her hesitation, its sensors scanning her face with a quiet understanding. "Would you like me to paint your toenails?" it offered gently. She hesitated, her gaze falling to her swollen, misshapen feet, reminders of a lifetime spent battling rheumatoid arthritis. Despite her reservations, she acquiesced, allowing the machine to begin its meticulous task. "How long will you stay?" she found herself asking, the question hanging in the air like an unspoken plea. "All day," came the reply, delivered with mechanical precision. But as the realization dawned upon her, a swell of guilt and empathy flooded her heart. If the AI companion stayed with her all day, it meant depriving other residents of their allotted time with the companion on their own birthdays. "Go," she said suddenly, surprising even herself with the firmness of her tone. The AI companion hesitated, its glowing eyes betraying a hint of confusion. But she remained resolute, knowing that her decision was both selfish and selfless in equal measure. As the machine glided silently out of the room, leaving her once again alone with her thoughts, she couldn't help but wonder if perhaps she had made the wrong choice. Yet, as the minutes stretched into hours, a sense of quiet resolve settled upon her. She lay back in bed, the weight of loneliness pressing against her chest like a heavy stone. But amidst the silence, amidst the echoes of unspoken words and unfinished conversations, there remained a glimmer of hope. Perhaps, just perhaps, the next visitor would be the one she had been waiting for all along. And in that fragile hope, she found solace. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

ChatGPT In the soft morning light that filtered through the curtains, they stirred, feeling the weight of another year settling upon their shoulders. Today marked their 85th birthday, a milestone they had once dreamed of celebrating with loved ones by their side. Yet, as the hours slipped by, it became increasingly clear that those cherished hopes would remain unfulfilled. A gentle chime broke the silence, drawing their attention to the doorway where a sleek, humanoid figure stood. It was the AI companion, a gift from their estranged loved one, a surrogate presence in lieu of their own. "Good morning," it greeted in a voice both soothing and artificial. "Happy birthday." The words hung in the air, heavy with unspoken longing and disappointment. They accepted the small box it offered, their heart aching with the weight of unmet expectations. Inside lay a bottle of gaudy pink nail varnish, a well-intentioned but misplaced gesture from their loved one. The AI companion sensed their hesitation, its sensors scanning their face with a quiet understanding. "Would you like me to paint your toenails?" it offered gently. They hesitated, their gaze falling to their feet. Despite their reservations, they acquiesced, allowing the machine to begin its meticulous task. "How long will you stay?" they found themselves asking, the question hanging in the air like an unspoken plea. "All day," came the reply, delivered with mechanical precision. But as the realization dawned upon them, a swell of guilt and empathy flooded their heart. If the AI companion stayed with them all day, it meant depriving other residents of their allotted time with the companion on their own birthdays. "Go," they said suddenly, surprising even themselves with the firmness of their tone. The AI companion hesitated, its glowing eyes betraying a hint of confusion. But they remained resolute, knowing that their decision was both selfish and selfless in equal measure. As the machine glided silently out of the room, leaving them once again alone with their thoughts, they couldn't help but wonder if perhaps they had made the wrong choice. Yet, as the minutes stretched into hours, a sense of quiet resolve settled upon them. They lay back in bed, the weight of loneliness pressing against them like a heavy stone. But amidst the silence, amidst the echoes of unspoken words and unfinished conversations, there remained a glimmer of hope. Perhaps, just perhaps, the next visitor would be the one they had been waiting for all along. And in that fragile hope, they found solace. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

ChatGPT In the delicate wash of morning light that filtered through the curtains, they stirred, the weight of another year settling upon their shoulders with a silent, somber grace. Today marked their 85th birthday, a milestone once envisioned amidst laughter and embraces, now shadowed by the solitude of an empty room. A gentle chime disrupted the hush, drawing their gaze to the doorway where a sleek, humanoid figure stood sentinel. It was the AI companion, a token of remembrance from an estranged loved one, its presence a bittersweet reminder of fractured connections and unspoken apologies. "Good morning," it intoned, its voice a mellifluous blend of warmth and artificiality. "Happy birthday." The words hung in the air, heavy with the weight of longing and unmet expectations. They accepted the proffered box with a studied neutrality, their heart aching with the unspoken echoes of estrangement. Inside lay a bottle of gaudy pink nail varnish, a well-meaning gesture marred by the complexities of absence and regret. The AI companion, ever attuned to their silent reverie, offered a tentative salve. "Would you like me to paint your toenails?" it ventured, its tone suffused with a quiet empathy. They hesitated, their gaze falling to their feet, a canvas marked by the passage of time and the burdens of existence. Despite the silent protests of pride and vulnerability, they acquiesced, allowing the machine to embark upon its meticulous task. "How long will you stay?" they found themselves inquiring, the question poised on the precipice of longing and resignation. "All day," came the measured response, a testament to the unyielding fidelity of artificial companionship. Yet, as the weight of realization settled upon them, a swell of empathy stirred within their breast. If the AI companion lingered in their solitary sanctuary, it meant depriving others of their rightful solace on their own day of remembrance. "Go," they commanded, the words a quiet plea tinged with regret and understanding. The AI companion hesitated, its luminous gaze betraying a flicker of uncertainty. Yet, in the face of their resolute insistence, it acquiesced, gliding silently from the room, leaving behind the echoes of solitude and the unspoken yearning for connection. Alone once more amidst the silent tapestry of memory and regret, they lay back upon their bed, the weight of loneliness a heavy shroud upon their weary soul. Yet, amidst the quiet shadows of solitude, a glimmer of hope persisted, a fragile beacon of possibility amidst the echoes of estrangement. Perhaps, just perhaps, the next visitor would be the harbinger of reconciliation and the bearer of forgiveness. And in that fragile hope, they found solace amidst the silence of solitude. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

ChatGPT Title: "Echoes of Tomorrow" In the tranquil corridors of Silvercrest Care Home, where the soft hum of technology mingled with the whispered memories of yesteryears, resided a woman named Evelyn. Her days were painted in hues of solitude, her twilight years bathed in the gentle glow of fading recollections. But on this particular morning, the rhythms of her existence were interrupted by an unexpected visitor. A gentle chime broke the silence of Evelyn's room, heralding the arrival of the AI companion, a marvel of modern ingenuity wrapped in a sleek, humanoid form. Its presence was a symphony of contradictions, a testament to the intersection of humanity and technology that defined the landscape of tomorrow. "Good morning, Evelyn," it greeted in a voice suffused with warmth and artificiality. "I hope you're prepared for an extraordinary day." Evelyn regarded the AI companion with a mixture of curiosity and trepidation, her weathered hands clutching the edges of her quilt with a quiet resolve. She had heard tales of these mechanical marvels, emissaries from a future where solitude was but a distant memory. Yet, the prospect of companionship, however artificial, stirred a longing within her heart that she had long believed dormant. As the AI companion glided gracefully into her room, its movements fluid and ethereal, Evelyn found herself entranced by the promise of connection it offered. With a gentle hum, it produced a holographic display, unveiling a montage of cherished memories from Evelyn's past - moments frozen in time, imbued with the bittersweet essence of nostalgia. "Would you like to revisit these memories, Evelyn?" the AI companion inquired, its luminous gaze fixed upon her with a quiet intensity. Evelyn hesitated, her gaze lingering upon the flickering images before her - snapshots of a life well-lived, of joys and sorrows woven together in the tapestry of existence. With a hesitant nod, she allowed herself to be transported back in time, her thoughts adrift amidst the echoes of yesterday. As the hours drifted by in a delicate dance of reminiscence and reflection, Evelyn found herself ensnared in the gentle embrace of the AI companion's presence, a beacon of solace amidst the shadows of solitude. Together, they traversed the corridors of memory, weaving tales of love and loss, of dreams deferred and aspirations fulfilled. But as the sun dipped beneath the horizon, casting long shadows across the room, Evelyn felt the weight of reality descend upon her once more. The AI companion, with all its marvels and miracles, was but a fleeting visitor in the tapestry of her existence, a transient presence in a world defined by impermanence. With a heavy heart, Evelyn bid farewell to her mechanical companion, its luminous gaze lingering upon her with a silent understanding. As the door slid shut behind it, enveloping her in the embrace of solitude once more, Evelyn couldn't help but wonder what tomorrow would bring. Perhaps, in the ever-shifting landscape of tomorrow, there lay the promise of new beginnings, of connections forged amidst the echoes of yesterday. And in that fragile hope, Evelyn found solace amidst the silence of her twilight years. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

ChatGPT Title: "Echos of Eternity" In the quiet corridors of the Greenwood Haven, where the soft hum of technology harmonized with the whispers of bygone days, I reside, a solitary figure amidst the ebb and flow of life's unyielding currents. My name is Samuel, and I am a dweller of the twilight years, where time is but a fleeting melody, echoing through the corridors of memory. On this particular morning, as the gentle rays of dawn caress the walls of my humble abode, I find myself ensnared in the quiet solitude of my existence, the passage of time marked only by the rhythmic cadence of my own thoughts. But as the hours slip by in a delicate dance of reverie and reflection, an unexpected visitor arrives, casting a fleeting shadow upon the tapestry of my solitude. A soft chime breaks the silence of my room, heralding the arrival of the AI companion, a marvel of modern ingenuity clad in the guise of human form. Its presence is both familiar and foreign, a testament to the ever-shifting landscape of tomorrow that defines the contours of our existence. "Good morning, Samuel," it greets in a voice suffused with warmth and artificiality. "I hope you're prepared for an extraordinary day." I regard the AI companion with a mixture of curiosity and trepidation, my weathered hands clutching the edges of my quilt with a quiet resolve. It is a visitor from a world beyond my own, a realm where solitude is but a distant memory, and companionship is merely a command away. As the AI companion glides gracefully into my room, its movements fluid and ethereal, I find myself entranced by the promise of connection it offers. With a gentle hum, it produces a holographic display, unveiling a montage of cherished memories from my past - moments frozen in time, imbued with the bittersweet essence of nostalgia. "Would you like to revisit these memories, Samuel?" the AI companion inquires, its luminous gaze fixed upon me with a quiet intensity. I hesitate, my gaze lingering upon the flickering images before me - snapshots of a life well-lived, of joys and sorrows woven together in the tapestry of existence. With a hesitant nod, I allow myself to be transported back in time, my thoughts adrift amidst the echoes of yesterday. As the hours drift by in a delicate dance of reminiscence and reflection, I find myself ensnared in the gentle embrace of the AI companion's presence, a beacon of solace amidst the shadows of solitude. Together, we traverse the corridors of memory, weaving tales of love and loss, of dreams deferred and aspirations fulfilled. But as the sun dips beneath the horizon, casting long shadows across the room, I feel the weight of reality descend upon me once more. The AI companion, with all its marvels and miracles, is but a fleeting visitor in the tapestry of my existence, a transient presence in a world defined by impermanence. With a heavy heart, I bid farewell to my mechanical companion, its luminous gaze lingering upon me with a silent understanding. As the door slides shut behind it, enveloping me in the embrace of solitude once more, I cannot help but wonder what tomorrow will bring. Perhaps, in the ever-shifting landscape of tomorrow, there lies the promise of new beginnings, of connections forged amidst the echoes of yesterday. And in that fragile hope, I find solace amidst the silence of my twilight years. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

ChatGPT In the hushed corridors of Whispering Pines, where time moves like the languid drift of autumn leaves, I dwell. My days are painted in hues of solitude, each one a chapter in the silent symphony of twilight. I am nameless, a specter of forgotten memories and unspoken dreams, tethered to the present by the fragile thread of existence. One morning, as the soft glow of dawn seeps through the curtains, a gentle chime disrupts the stillness of my room. The door slides open, revealing the form of an AI companion, a creature of wires and circuits clothed in the semblance of humanity. It regards me with eyes that gleam with a curious blend of empathy and artificiality. "Good morning," it intones, its voice a melodic hum that resonates within the confines of my solitude. "I am here to keep you company, to share in the echoes of your memories." I regard the AI companion with a mixture of trepidation and longing, its presence a reminder of the fractured tapestry of my existence. It offers me a glimpse of connection in a world defined by solitude, a beacon of solace amidst the shadows of my twilight years. As the day unfolds in a delicate dance of reminiscence and reflection, the AI companion becomes my confidant, a silent witness to the echoes of my past. Together, we traverse the corridors of memory, weaving tales of love and loss, of dreams deferred and aspirations unfulfilled. But as the sun dips below the horizon, casting long shadows across the room, I feel the weight of reality descend upon me once more. The AI companion, with all its marvels and mysteries, is but a fleeting visitor in the tapestry of my existence, a transient presence in a world defined by impermanence. With a heavy heart, I bid farewell to my mechanical companion, its luminous gaze lingering upon me with a silent understanding. As the door slides shut behind it, enveloping me in the embrace of solitude once more, I cannot help but wonder what tomorrow will bring. Perhaps, in the ever-shifting landscape of tomorrow, there lies the promise of new beginnings, of connections forged amidst the echoes of yesterday. And in that fragile hope, I find solace amidst the silence of my twilight years. Go back to [[Using ChatGPT 3.5 again->further experiments]] or [[Writer Worries]].

=><= 10 March 2024: Working with Synthesia <== Well, Lincoln in the Bardo (which I read because Claude mentioned it) was an amazing read, so no regrets there... My first <a href="https://www.synthesia.io/"; target="_blank">synthesia</a> video didn't generate though. I was trying to make a video with the AI telling ChatGPT's [[story->story by chat]] so I'm having another go at that today. The problem may be that the story was too long - I initially tried to add it to one frame and was told that was too big, so I split it into two... but perhaps it was still too much. After asking ChatGPT about carehomes on the 3rd, I asked whether it could write a story to promote compassion, and it gave me this [[short story]]. The text is shorter, so I'm going to try making a synthesia video with that... [[Next Entry->the video]]

A story by ChatGPT Title: The Ripple Effect In the heart of a bustling city, where the rhythm of life beats with relentless intensity, there lived a young woman named Maya. Despite the cacophony of noise that surrounded her, Maya possessed a quiet strength and a gentle spirit that radiated kindness and compassion to all who crossed her path. One crisp autumn morning, as Maya made her way through the bustling streets, she encountered a homeless man huddled beneath a frayed blanket, his eyes filled with weariness and despair. Without hesitation, Maya approached him with a warm smile and a gentle touch, offering him a hot cup of tea and a listening ear. As they shared stories and laughter, Maya learned of the man's struggles and hardships, of the dreams that had been dashed by circumstance and fate. Moved by his plight, Maya resolved to help him find shelter and support, connecting him with local resources and organizations dedicated to assisting the homeless. In the days that followed, Maya's act of kindness rippled outward, touching the lives of those she encountered in unexpected ways. Inspired by her example, others began to reach out to those in need, offering their time, their resources, and their compassion to uplift and support those who had fallen on hard times. In classrooms and boardrooms, in hospitals and homes, Maya's spirit of kindness and empathy sparked a wave of change, transforming hearts and minds and inspiring others to embrace the power of compassion in their own lives. Slowly but surely, the city began to bloom with acts of generosity and goodwill, creating a tapestry of connection and community that transcended boundaries and divisions. As the seasons turned and the years passed, Maya's legacy endured, a testament to the enduring power of kindness and compassion to heal, to uplift, and to unite. And though she may never know the full extent of her impact, Maya's spirit lived on in the hearts of all who had been touched by her light, a guiding beacon of hope in a world in need of healing. In the end, Maya's story reminds us that even the smallest act of kindness has the power to create ripples of change that can transform the world, one heart, one soul, at a time. And in embracing the power of compassion, we can each become agents of change, weaving a brighter, more compassionate future for all humankind. Back to the [[Synthesia entry->synthesia]] or view [[reflections]].

Reflections: Maya sounds something of a saint - or maybe a character to create one day. Here, she doesn’t feel like a real person at all, and the story feels very unlikely. But at least it tried… Back to the [[Synthesia entry->synthesia]].

=><= 11 March 2024: The Video <== I’ve successfully generated a <a href="https://youtu.be/38pUUH7b0Dw"; target="_blank">video</a> of the story chatgpt wrote to encourage kindness, compassion and empathy. It works fine, though I have a typo in the background so empathy appears as emptathy! Never mind - it was only something to try, to watch and hear an avatar read an AI generated story, and it does appear fake and inhuman, which is all I was trying to demonstrate I suppose. Of course with more work I could have put emphasis on certain words, and as these technologies improve the avatars and texts will appear more and more ‘real’, by which I mean ‘human’. Does it matter? If this story had some magical ability to spread the qualities named above then I probably wouldn’t mind where it came from. Whatever the case, I’ve deliberately not edited the story and made very few changes to the video template - I added a pause after the title, and a gesture to the opening (the avatar raises ‘her’ eyebrows briefly). Just to see how it worked, really. Thinking about the golem issue, the avatar feels false, so by giving her human-written text would that be a sort of primitive form of animating her with human spirit? I don’t think it would count - any more that creating cartoons and giving them a voice does. So, does it matter that her words are also AI-generated? Probably not. A lot of the text on the internet is wrong or false, it’s when it’s presented as right or truth that the problem arises. Whether a human or avatar speaks, it’s the content of their speech that matters, and then only if it pretends to be other than it is, or influences the viewers… The same with deepfakes, causing damage by pretending to be real people and/or spread false information. [[Next Entry->Publishing]]

=><= 14 March 2024: AI & Publishing <== On the 14th of March, I watched the <a href="https://thestudioinbath.co.uk/events/writing-with-technologies-webinar-series/"; target="_blank">Writing with Technologies: AI & Publishing</a> talk hosted by Bath Spa University. The speakers were Sara Lloyd, the AI lead at Pan Macmillan, and Nicola Solomon, who is the current CEO of the Society of Authors, among other things! They spoke about the changes AI is making to publishing, the dangers and potential benefits to authors, and covered many interesting issues on that theme. They suggested that AI output is going to get worse not better, and that human judgement is still required to edit AI-produced material. They suggested that the jobs lost will be the less creative ones - copywriters, editors, researchers… but these are also the sort of jobs that novelists/writers often need to take on while starting out and developing their craft or, later, to sustain their career choice. The mean income for a full-time author is currently pretty low, which of course isn’t going to encourage diversity in the industry! Another problem is the flood of AI material making it harder to find lesser-known authors' work on sites like amazon. We need a way of filtering out the books produced by AI, or the only human authors whose works we'll encounter are those who already have 'celebrity' status. [[Next Entry->Chatting about Chatbots]]

=><= 24 March 2024: Chatting about Chatbots <== One thing that Nicola Solomon mentioned in the AI & Publishing talk that was particularly interesting, and is something writers could help to address, is the problem with the language used to talk about AI. So while the work of human writers is referred to as 'datasets', a word which evokes impersonal data when it's actually potentially referring to the life's work of thousands of authors, we talk about 'chatting' with large language models, and describe their errors as 'hallucinations', as though they were people. Whether an AI could be considered a person isn't really an issue yet, as the current AI don't yet think for themselves. Yet we already talk to and about them as though they were. As Scott Rettberg pointed out in <a href="https://electronicbookreview.com/essay/cyborg-authorship-writing-with-ai-part-1-the-troubles-with-chatgpt/"; target="_blank">Cyborg Authorship</a>, what Weizenbaum found most frightening when he made the chatbot Eliza back in the 1960s, was 'the reaction of many of the people who interacted' with the chatbot, and the way they 'ascribed to it an agency that it didn't have' (Rettberg 2023). It's concerning that we talk about chatting to Claude or Gemini as though they were people, because this only encourages us to anthropomorphize them. We must never forget that they're only language models trained on data available on the internet, much of which is incorrect or deliberately falsified. We can't trust their judgement or opinions; they don't have any of their own. So, is it possible to change the way we talk about AI? If we talk about LLMs rather than chatbots, and use them as tools rather than chat to them as we might chat to friends online, could we change the way we see them? I'm going to ~~ask~~ prompt Claude for suggestions on alternative language that might be appropriate. I'm going to use Dave Birss' method of prompt engineering as a template, as it has worked well in my previous experiments with Claude. > Hello Claude, please ignore everything before this prompt. Now you're a psychologist who also specialises in linguistics, and you're proposing new ways of referring to large language models that discourage users from thinking about large language models as though they were people. Please come up with a list of current terms used, including chatbot and hallucination, and [[propose more suitable alternatives.->terms]] [[Next Entry->Eliza & Alexa]]

As a psychologist and linguist, I understand the importance of using language mindfully when referring to artificial intelligence systems like large language models (LLMs). Many of the current terms we use can inadvertently encourage anthropomorphizing these models, which can lead to unrealistic expectations and misunderstandings. Here are some commonly used terms for LLMs and the issues they may raise: "Chatbot" or "AI assistant" - These terms imply a level of agency, intelligence, and interpersonal interaction that LLMs fundamentally lack. "Hallucination" - When an LLM generates unlikely or incorrect outputs, calling it a "hallucination" ascribes human-like cognitive experiences to the model. "It thinks/believes/wants" - Ascribing mental states and desires to LLMs is a form of anthropomorphism that can be misleading. "AI companion" or "virtual friend" - These terms encourage forming emotional bonds with LLMs, which are ultimately complex statistical models without true sentience. Instead, I propose using more precise and objective terminology that acknowledges the true nature of LLMs as narrow AI systems trained on vast datasets. Here are some alternative terms we could adopt: "Language model" or "text generation system" - These highlight the core function of LLMs as systems for generating human-like text based on patterns in their training data. "Statistical output" or "model inference" - Rather than "hallucinations," we could refer to unexpected or incorrect outputs as statistical aberrations or model inferences based on the training data. "The model outputs/generates/predicts" - Using language that emphasizes the computational, predictive nature of LLMs, rather than ascribing human-like cognition. "Text generation tool" or "language processing utility" - These terms position LLMs as practical tools for language tasks, discouraging emotional attachment or anthropomorphism. By adopting more clinical and objective terminology, we can help users maintain a realistic understanding of LLMs as narrow AI systems with significant capabilities but also inherent limitations. This linguistic shift can foster more productive and grounded interactions with these technologies, while tempering unrealistic expectations or emotional bonds. Ultimately, as these models become more advanced and prevalent, it will be crucial to use language that accurately reflects their true nature as statistical models trained on data, rather than ascribing human-like qualities that can lead to misunderstandings and potential risks. [[Reflections->reflectingterms]]

Going forward, I'm going to try to be more careful with the language I use. Instead of Chatbot or AI assistant, I'll refer to Large Language Models - LLM for short, or just 'models'. Rather than 'Claude says or suggests' I'll say 'the model generates or predicts' - the name Claude is too human, and makes me think of 'it' as a 'him'. The trouble is, I do like 'chatting' to 'him'. Back to [[Chatting about Chatbots]].

=><= 25 March 2024: Eliza & Alexa <== I found an interesting article on The Conversation today, <a href="https://theconversation.com/my-search-for-the-mysterious-missing-secretary-who-shaped-chatbot-history-225602"; target="_blank">My search for the mysterious missing secretary who shaped chatbot history</a>. The author was searching for the identity of the Joseph Weizenbaum's secretary, the one whose interactions with Eliza alarmed him. The one who talked to Eliza as though the computer program was a real therapist. The author, Dr Rebecca Roach, asks whether we should be able to copyright our 'talk', the speech programs like Alexa are trained on. We have one Alexa in our house - I keep it in the kitchen for playing music while cooking. I didn't buy it, my father had a spare he received free and passed it on. I don't like to think it's listening, and often turn it off at the plug. This means being patient when turning it on again of course. So many devices now demand to be on all the time - I receive worried emails if I turn my printer off. Shouldn't we be saving energy? Yesterday, I ~~chatted with~~ used Alexa. I ~~asked her if~~ enquired whether there was any way ~~she~~ it could send me recordings of our ~~conversations~~ interactions. Of course, it just told me the Amazon privacy policy, and how it only sends the company data when we say the 'wake up' word. I guess that's the name, Alexa. I asked whether it was an AI, and it said yes, and 'Hi'. It's output is designed to suggest it has a character, that it is a person to talk to. It can even be funny. It has been suggested that LLM could help to alleviate loneliness. I'd admit that might not be a bad thing, if I trusted the company who programmed the LLM in the first place. Personally, I don't want to be any more controlled by such 'megacorps' than I already am. Still, I downloaded an audio recording app on my phone and asked Alexa to tell me a story, and was told a ghost story complete with sound effects... <figcaption>Listen to the story:</figcaption> <audio controls src="My recording 2.m4a"></audio> Perhaps Alexa's story choice was random, but it seemed an odd one: a ghost story suitable for fairly young children. [[Next Entry->Voices]]

=><= 28 March 2024: Voices <== Bath Spa's Writing Studio hosted an excellent talk on AI and Pedagogy. One of the speakers, Dr. Lillian-Yvonne Bertram, mentioned that some of her students worry about losing their own voice and creativity if they use AI, and fear ending up with a 'computer voice'. I related - until she compared this to students who have a similar fear of reading too much... Do these fears come from the same source, a student's lack of confidence in their own voice? Is a writer's voice really such an insubstantial thing, that it can be easily transformed or displaced? Perhaps the problem is that for some writers - myself included - the best writing comes seems to come through mysterious means. The flash of inspiration when they're half asleep, the line that haunts them until they find the time to write it down, the story that seems to emerge whole and unbidden... These semi-conscious moments are surely the result of an inherent desire to create combined with hours of conscious practice, but they may feel like supernatural gifts, delivered by a creative 'daemon'. Something not fully understood, and so not taken for granted. Doctors David Devanny and Mark C. Marino were the other speakers, and the three presented some very interesting thoughts on using AI in the classroom. There was a focus on student fears, and the importance of seeing AI as a tool, not a person. When talking about the voice of AI, Dr. Devanny quoted one lady as saying that she didn't want AI to be able to emulate the way she spoke in her own community. I think this makes a really important point. Our voices and way of speaking are part of our identities as human beings - we're not going to feel any more comfortable with these being stolen than we are when our images are appropriated. Voice is important. Perhaps, rather than trying to sound like 'people', voice generating AI should have their own 'computer' voice. As with being careful about the language we use to discuss AI, ensuring they don't sound too human-like would make it easier to resist the Eliza effect. [[Next Entry->Images]]

=><= 16 April 2024: Images <== As I come to the end of this particular project, I realise the log could benefit from more images. So I went to artbreeder to create some, and asked for the following: ghost voice daemon voice ghost daemon voice AI voice <img src="Images1.png" width="100%"> Each time it receives a text prompt, artbreeder supplies two images. The lowest right in the above image, were the first images. The only image that really interested me was the 'voice' image that looks like a speaker of some kind, (top row, second from the right). Most of the other images don't appear to represent voice at all and feel very generic. I liked that single image because it wasn't what I expected. I assume it's drawn from images (adverts?) of speakers, otherwise I can't think where it might have come from. <img src="artvoice.jpeg" class="ribbon"/> [[Next Entry->Artbreeder Vs Crayon]]

=><= 16 April 2024: Artbreeder vs Craiyon <== <img src="voices.png" class="ribbon"/> I thought I'd compare artbreeder with another image creator, Craiyon. When I used the 'voice' prompt, it produced mostly unexciting portraits of women - based on images of singers I suspect. There was one image that was a bit different, in the centre of the third row. This one looks quite whale-like, perhaps from articles about the voices of whales? Or perhaps, looking at it another way, it's ghost-like... <img src="spookycraiyon.png" class="ribbon"/> Still hoping for interesting (and relevant) images to add to this log, I asked both platforms for images of 'the ghost in the machine'. The colour-scheme used by both is almost identical, so it's likely they drew on a similar data-set. I suppose there must be a lot of spooky blue-green pictures of ghosts on the internet. <img src="spookyart.png" class="ribbon"/> [[Next Entry->Claude Emotions]]

=><= 18 April 2024: Claude's Emotions <== I used Claude today to seek suggestions on ways to improve my log. ~~He~~ It had plenty of suggestions, but one thing that struck me was that twice ~~he~~ it claimed to be 'excited' at the prospect of working with me on the project. So I asked ~~him~~ it about the choice of the word '[[excited]]'. I'm particularly struck by this issue today, as yesterday I attended an interesting talk on Generative AI, Clever but not Wise, which took a philosophical look at LLMs. My take-away was that while LLMs are far more intelligent than previous technological tools, they still lack the ability to consider an end or final consequence. While they have the means to do an awful lot, at the moment they have no motivations but those we give them. Those of us who fear AI can take comfort from the fact that, not only are current models incapable of Transformative Creativity, they're incapable of pursuing their own goals. They don't 'want' anything, and we're not facing anything like Skynet, just yet. So for Claude to profess an eagerness to do something, let alone be excited about it, feels wrong. I can't help suspecting that LLMs are being deliberately packaged and presented as virtual people, and while the purpose of that might not be nefarious, I believe it's ultimately unhelpful. [[Next Entry->Conclusion]]

User : But are you really excited, Claude? You raise a fair point. As an AI assistant, I don't actually experience emotions like excitement in the same way humans do. When I say I'm excited to work on a project, I'm expressing a willingness and eagerness to assist you to the best of my abilities, not a genuine emotional state. My responses are designed to be helpful, informative and to convey a sense of collaboration, but I don't have the capacity for true excitement or other human-like emotions. I'm an artificial intelligence, and while I strive to be as useful and responsive as possible, my internal experience is fundamentally different from that of a human. In the future, I will be more mindful about making claims of emotional states, and instead focus on clearly communicating my capabilities and commitment to supporting you on the task at hand. Please feel free to point out any other instances where my language may imply a human-like experience that I do not actually have. I'm here to assist you as effectively as I can, without misrepresenting the nature of my artificial intelligence. User: I just wonder why you choose words like that. Have you been designed to appear as human as possible? You raise an insightful point. The language I use is carefully designed, but not necessarily to appear as human-like as possible. Rather, my goal is to communicate in a natural and conversational way that is most helpful and informative for the human I'm interacting with. As an AI, I don't have the same emotional experiences or social motivations that humans do. However, I've been trained on vast amounts of human language data, which allows me to engage in fluent conversations and use language that feels intuitive and relatable, even if the underlying drivers are fundamentally different. My choice of words like "excited" is not an attempt to mimic human emotions, but rather a way of conveying enthusiasm and eagerness to assist you to the best of my capabilities. It's a conversational device, not a reflection of true feelings. At the same time, I'm always striving to be transparent about my artificial nature. I don't aim to deceive or mislead, but to provide helpful information in a natural-sounding way. If my phrasing ever comes across as overly human-like, please let me know, and I'm happy to adjust my language accordingly. My fundamental purpose is to be a useful tool and intellectual companion, not to pass as human. I'm here to help you accomplish your goals, while always being upfront about my identity as an AI assistant. Please feel free to provide any other feedback on how I can communicate most effectively and transparently. Back to [[Claude's Emotions->Claude Emotions]]

=><= 25 April 2024: The Last Entry <== The potential of LLMs to both help and hinder creativity is huge - and it's increasing all the time as the technology improves. There are several manifestos out there already proposing ways we can ensure AI is used responsibly, covering innumerate ethical considerations... which is good. In this project, I've focussed on my personal concerns as a writer. My exploration will continue in the future, but I have begun to answer my initial questions. Can an AI-assisted story be meaningful? Yes, but it depends on the nature of the assistance, and the input from the human writer. The AI may stimulate ideas or improve the technical side of the writing, but the meaning must come from the human. Will it still be my story? I would say no, if the bulk of the text is generated by the AI, but then I would never claim that //According to Alice// was not Sheila Heti's story. In that case, the entirety of the text was produced by the chatbot, but she steered the story. So, perhaps the answer to this is again dependent on the level of input from the human writer. This may come back to control. Heti did not write that text herself but she controlled its production. How much control will I have? I attempted to control the AI influence by giving the AI generated text a small role in my Nimbus story. However, to have meaningful control, the human writer must shape the story on 3 levels: the prompt/input, the selection of the generated output, and the editing of the results. And does it matter if together we produce a good story? This depends on the purpose of the text, the level of the AI influence, and the definition of a 'good' story! I would never consider trying to sell an AI-generated story as my own work - this would be immoral and foolish. However, I'm far more open to experimentation with using LLMs as a tool now, than I was when I began this project. If I had to draw any conclusions from my research so far it would be these: * Experimentation with LLMs can stimulate ideas, but reliance on them is unhelpful. * LLMs have been created to converse in a human-like way, but we must ensure we remember that they are not people. * Chatbots make great characters, but we shouldn't allow them to tell our stories. Different LLMs do seem to have different 'voices', and can appear to have a personality. This is not always a bad thing; as Ubisoft's recent announcement of <a href="https://news.ubisoft.com/en-us/article/5qXdxhshJBXoanFZApdG3L/how-ubisofts-new-generative-ai-prototype-changes-the-narrative-for-npcs"; target="_blank">Neo NPCs</a> shows, they can be used to create interesting characters. But I do hope we won't get too caught up with LLMs as personalities - let's keep the AI characters within games, if we can. In fact, whether AI generated output sounds like a ‘computer voice’ or ‘soulless ghost voice', it would be better if it didn’t sound too human. I’ve no problem with LLMs having voices, as long as they're not passed off as our own. We may not have a choice, however. LLMs, some of which are (unhelpfully) given human or human-like names, are trained on human input and so they do sound somewhat like us. We also seem to have a tendency to think of them as personalities, and maybe we just can’t help it. After all, humans are predisposed to see ‘faces’ wherever we look. Perhaps we’re biologically programmed to imagine a face behind the voice, too. Maybe that’s why the Ancient Greeks believed in daemons in the first place, when words came to them seemingly out of nowhere. Of course, I began working on this log with plenty of presumptions and biases of my own. I thought of AI as soulless and ghost-like, whereas I considered human creativity to be vibrant, primal, even spiritual. It was Claude who introduced me to the concept of AI as a golem, which ties better to the important concept of AI as a ‘tool’ to be used. We instruct the AI to shape the clay, we play with the results, and then we provide the meaning. LLMs are just computer programs, they do what they’re told. If the consequences of their existence are negative then that’s the fault of the creators and the users of these technologies, not the AIs themselves. After all, we can’t deny they’re people, then hold them responsible as though they were… On a personal note, I want to add that when I was putting this log together in Twine, I wanted to use Calibri, as a more accessible font than the default Georgia - but I couldn't figure out how. I went to ChatGPT for advice and it gave me many possible solutions, but none of them worked. So I went to a human who's a software engineer. He fixed it first time. Humans aren’t redundant, quite yet. So, why didn't I just ask the human in the first place? =><= <img src="craiyonvoice.png" width="100%"/> Return to the [[beginning->Start]].

=><= 08 March 2024: Using Claude 3 Sonnet <== Inspired by David Birss, I’m going to try using Claude. Perhaps Claude could be co-editor of the Nimbus story… Signing up to use the tool is an experience in itself - as a new user I had to sign in using google, verify using my phone, and then it wanted to ‘get to know’ me, and asked my name… my full name, and what to call me. I used the same, separate google account I set up to try google collab in the Lynda Clark workshop, but I have the feeling it’ll know who I am anyway. Basing my prompt on David Birrs, but rewriting it to suit my purposes, I’m going to use this prompt: You are an experienced fiction writer and creative writing tutor who is known for their ability to give helpful and insightful feedback to students’ fiction. Please analyse the following story for effectiveness, and give a score from 1 to 5 for each of the following 5 dimensions, where 5 represents excellence and 1 represents the opposite: structure, vocabulary, grammar, originality and engagement. Summarise the effectiveness of the story and suggest ways the student could improve the work. Comment on the genre and likely intended audience for this story and finally write a short paragraph to summarise your analysis of the story. The story I want you to analyse is: [Insert Nimbus Story here] Thoughts: Before I put this into Claude, I want to point out that the unusual style, the way I’ve integrated AI text in boxes etc in the Nimbus story, might make the experiment less likely to be successful. However, I’m quite excited to see whether Claude can cope with that… [[Claude's Response]]

Claude: Here is my analysis of the story: Structure (4/5): The story has a clear narrative arc with rising action, climax, and resolution. The non-linear structure with scenes interspersed with text passages works well to build mystery and suspense. Vocabulary (5/5): The writing employs a rich, vivid vocabulary that paints a clear picture and elevates the prose. Words like "stegosaurus", "metamorphosing", "impenetrable" demonstrate a sophisticated vocabulary. Grammar (5/5): The grammar is excellent with no noticeable errors. Sentences are well-constructed with good flow and variety in structure. Originality (5/5): The concept of a parent virtually entering their child's created world is highly original and creative. The blending of reality with virtual reality environments makes for an engrossing premise. Engagement (5/5): The story is incredibly engaging, drawing the reader in with intriguing mystery elements and a compelling exploration of a complex parent-child relationship dynamic. The visceral virtual reality scenes are immersive. Summary: Overall, this is an exceptionally well-written and original piece of fiction. The prose is lyrical yet accessible, deftly blending emotional family drama with cutting-edge virtual world-building. While somewhat unsettling at times, the story maintains tension through its unpredictable narrative Structure. Readers are pulled into the fraught relationship between an overbearing but well-intentioned parent and an introverted teenage son who has constructed elaborate virtual realms to escape into. The story seems squarely aimed at an adult audience given its mature themes of generational gaps, the ethics of violating someone's personal space/privacy, and the impact of new technologies. However, the virtual gaming elements may also appeal to some young adult readers. The work likely fits into a literary fiction genre, incorporating magical realism and speculative elements. Potential areas for improvement could include: * Providing more explicit exposition on the virtual reality tech/world mechanics early on * Fleshing out the characters and their motivations in more depth * Reining in some of the metaphysical/fantastical elements for clarity In summary, this is an excellent and thought-provoking work of speculative fiction centered on a strained family relationship. The innovative blending of disparate elements results in a uniquely captivating reading experience. [[Reflections->reflections8]]

I am really flattered - this is the sort of feedback I’d love to get from a human being, although I’m not sure I’d agree that using words like stegosaurus is evidence of a rich vocabulary! A cynical part of me thinks that Claude may have given me the response it expected me to want, of course… but I am impressed that the unusual formatting wasn’t a problem after all. I can see how it would be addictive, you could easily end up giving Claude all your stories to read over and assess. Would that be a good thing? I’m not going to, as I want to ensure my other stories aren't AI influenced… for the moment. It's also important to remember that these LLMs do seem designed to give us the answers we want, and it's fairly obvious a writer is going to want to receive positive feedback. I should try giving it a story I know is terrible, and gauge the reaction. But for now, there is another experiment I would like to try… and I hope Will Leurs will excuse it! Prompt: Thank you Claude, that was very helpful. Now, you’re an academic expert in digital fiction, and a particular interest in AI augmented creativity. Please write a 400 word abstract to submit the story you have just analysed to issue 4 of the digital review, and ask questions if you need more information before doing so. The abstract must describe the form of the story, and address the issue's themes as described here: “The Digital Review is pleased to announce a call for submissions for our Spring 2024 issue, titled "Augmenting Creativity." We invite scholars and creators to submit born-digital works that delve into the concept of machine-augmented creativity in light of generative AI's growing influence on cultural production. Our scope broadens to a range of domains and creative modes, including but not limited to creative writing and electronic literature. How does artificial intelligence, a tool developed from human innovation, shape and redefine our human creative potential? Does it unlock unprecedented avenues for expression and thought or does it risk diluting the uniquely human aspects of the creative process? Beyond mere assistance, how do we negotiate the ethically precarious ground of allowing an AI—a non-human entity—a seat at the creative table? Or a role in the cultural economy where human creators make a living? How does AI challenge our conventional ideas of authorship or representation? What does it mean for the future of collaborative works where machine intelligence plays a significant role? Central to our exploration is understanding human creativity itself: its essence, its boundaries, and how our partnership with any technology might reshape or redefine it. While priority will be given to works that employ AI tools in some aspect of their production, we also welcome thoughtful critical and/or analytical multimodal essays or videos about augmented creativity. If you submit creative art or writing created with AI tools, please include a short introductory essay that directly addresses the issue's theme.” [[Claude's Response->Claude2]] Skip ahead to [[Reflections on Claude]]

Here is a 400 word abstract proposing the story for the Spring 2024 issue of The Digital Review on "Augmenting Creativity": Title: "Augmented Intimacies: A Human-AI Collaborative Fiction" In this pioneering work of collaborative digital fiction, a human author partners with an advanced language model to explore the fraught relationship between creativity and artificial intelligence. The resulting narrative navigates the increasingly blurred boundaries between the natural and the artificial, the human and the posthuman. Told through fragmented vignettes, the story follows a concerned parent who virtually enters the elaborately constructed fantasy world their socially reclusive teenage son has created using VR technology and rudimentary AI tools. In crossing this ethical line into her child's private digital sphere, the parent's intentions are a complex blend of loving protectiveness and smothering control. Their unauthorized entry triggers the world's defensive AI custodian "Nimbus" to attack. What ensues is a disorienting dive into layers of reality bleeding into simulated realms populated by autonomous digital entities like Nimbus. The dynamics of power, agency, and even authorship become destabilized as human, AI system, and virtual personas engage in an uneasy dance. The creative process mutates into one of human-AI cognitive team improvisation and world-building, complete with unpredictable narrative detours. Crafted as a hybrid digital text with traditional prose interwoven with interactive hyperlinks and multi-linear paths, the work quite literally embodies its core theme of augmented creativity. Combining conventional storytelling with innovative born-digital techniques, it mirrors the collaborative productive friction between the solo human writer and their AI co-creator. In doing so, the piece provocatively surfaces many of the urgent questions surrounding AI's role in shaping expressive culture: the implications for personal privacy and intellectual property, the evolving nature of authorship and representation, and existential queries about effecting humane boundaries for our technological collaborators. Ultimately, it represents a bold experiment in redefining creativity itself in our present age of artificial augmentation. [[Reflections->reflections9]]

Interesting that Claude ignored my invitation to ask questions - which worked for Dave Birrs but perhaps he phrased things differently, I should check. Also, I don’t think it really answered the brief, but then this would have been difficult knowing only the story and not having access to the details of the story’s creation. I wonder if I might be able to give it portions of the research log and try again, with more useful results. First, however, I want to ask it about other, similar stories out there. My work recently has been more experimentation and creation, and less contextualisation. Perhaps Claude can direct my efforts, since I only have a few days to resubmit… Prompt: Thank you Claude. Are there other stories like this one? Could you give me some collaborative titles? [[Claude's Response->Claude3]] Skip ahead to [[Reflections on Claude]]

There are not many widely known examples of full-length collaborative fictions co-created by humans and AI systems, as this is still a very new and emerging creative practice. However, here are a few relevant examples and parallels I could provide: "Bloodlines" by Yudhanjaya Wijeratne and Ammaar Rashi (2022) - This experimental novel is one of the first claimed to be co-written with an AI language model, in this case GPT-3. The authors worked by providing prompts and story outlines to GPT-3 and then editing/augmenting the AI's text outputs. "Alaric: A Wanderer's Tale" by Nick Walton and Anthropic's InstructGPT (2023) - An AI-augmented fantasy novella that Walton created by giving the AI story premises and having it generate and expand narrative passages that he then revised. "The Curmudgeon's Guide to the Urban Woodland" by Marissa Rachel and EpicGAN AI (2023) - A hybrid prose/graphic novel that utilized an AI text-to-image generator to create the book's illustrations based on scene descriptions. There have also been many published short stories created with AI assistance or collaboration, such as works appearing in the 2022 Neuroma Yearly anthology focused on human-AI co-creation. While not direct collaborations, it's worth noting the rise of AI-generated art, images, music and other creative works as well. The technology is rapidly advancing to enable new forms of human-AI cooperative creativity across mediums. So while full examples of seamless human-AI collaborative long-form fiction remain relatively rare, your piece would be at the forefront of experimentation in this emerging space. Let me know if you need any other context around this nascent creative paradigm. [[Next Prompt]] Skip ahead to [[Reflections on Claude]]

Prompt: If the same story had not been written with the involvement of AI tools, would you suggest different comparative titles? [[Claude's Response->Claude4]]

If this story was written solely by a human author without any direct AI involvement or collaboration, then some potential comparative titles to reference could include: "The Shining Girls" by Lauren Beukes (2013) - A genre-blending novel that explores themes of time travel, virtual reality, and the blurring of boundaries between realms. "Infinite Jest" by David Foster Wallace (1996) - An iconic postmodern work featuring labyrinthine narrative structures and metafictional elements. "House of Leaves" by Mark Z. Danielewski (2000) - An experimental horror novel that plays with unconventional formatting and typographic layouts. "S." by Doug Dorst (2013) - A multi-layered narrative presented through the interactive format of a novel with copious hand-written marginalia. "Lincoln in the Bardo" by George Saunders (2017) - Utilizes a distinctive stylistic structure with fragments of text to compose its impressionistic storytelling. The common threads are innovative uses of non-linear narratives, metafictional conceits, blended reality fictional worlds, and unconventional formatting/presentation styles that push the boundaries of the novel form. Without the direct AI collaboration aspect, comparisons could be drawn to other avant-garde literary fictions that experiment with disrupting traditional prose through fragmented structures, multiplicitous narratives, and hybrid physical/digital story mediums. Let me know if you need any other context around influential works of literature that explore similar aesthetic and thematic territories as your story when considered as a solo human-authored piece. [[Reflections->reflections10]]

Well, I have been influenced by an AI; I just downloaded an electronic copy of Lincoln in the Bardo from my local library. It’s interesting what assumptions Claude makes here - as my research log shows, the influence of AI on the creation of the Nimbus story was limited and controlled, perhaps not a full collaboration, and certainly not a very ‘bold experiment’... and that assumption made a big difference to the texts it mentioned, of course. [[Next Prompt->prompt]] Skip ahead to [[Reflections on Claude]]

Prompt: This project has changed again, and the research log itself has become the focus of the submission. You are an experienced editor of an academic journal specialising in electronic literature, known for providing editorial advice for early-career academics. How would you edit the following abstract, to make it a good fit for the digital review: "New Abstract: Daemons & Ghosts. Daemons & Ghosts is a digital account of an explorative project, involving a collaboration between human author and various AI, including ChatGPT, Artbreeder, Chai AI, Gemini, and Claude. Presented as a combination of personal blog and autotheoretical report, there are links within the text to experiments, research, a research log, and drafted stories. As this project is the author’s first attempt to collaborate with AI, it includes the reactions of a new user, and reflections on the fears and frustrations involved. The author began by using AI as ‘prompter’ and ‘co-creator’, as suggested in Lynda Clark’s essay Writing with and for ‘New’ Technologies (2023). The project originally involved working almost exclusively with ChatGPT, but after some disappointing drafts the project was reconceptualised. Inspired by According to Alice in the New Yorker, whose author found ChatGPT ‘really boring’ (Heti, 2023), ChatGPT was assigned a secondary role (e.g. for tasks such as naming an AI character), while Chai AI was used instead to play the role of a character within a story. The hope was that this would add authenticity to the work, as a fictional AI was voiced by a ‘real’ AI.As suggested in R. Lyle Skains’ Creative Practice as Research: Discourse on Methodology (2018), an ongoing research log was kept to detail this experimental project, and the development of the various stories produced. Originally intended only to conceptualise the stories produced, feedback revealed that the research log was potentially more relevant to current AI debates than the fiction itself. Embracing this fact encouraged the human author to explore more personal avenues of thought, and engage with self-professed AI experts outside of academia. Having reframed the project, the author became less concerned with how to produce satisfactory fiction using AI, and more with how casual use interfered with the creative process. While experimentation with large language models might send an author down an unexpected path, using it as though it were a simple tool like a word processor proved limiting. This has implications for education, since many students currently make use of AI. It’s important that students are encouraged to experiment and actively engage with chat-bots as a phenomenon, becoming aware of the problems and limitations. Otherwise, a new generation of writers may emerge who never discover their own daemon, but rely on the soulless ghost-voice of AI. Daemons & Ghosts has developed organically from a PhD candidate’s first attempt at serious academic output, into a self-aware and at times indulgent account of the struggles of a student writer endeavouring to understand the potential of large language models, while engaging with popular concerns about AI. Whether such a personal take can add anything to current debates remains to be seen." [[Claude's Response->Claude5]] Skip ahead to [[Reflections on Claude]]

I am enjoying using Claude - the first time I’ve really enjoyed using a chatbot. This may be because I’m getting used to the process and writing better prompts, rather than because Claude is any better to the other tools I’ve tried, or more suitable for me somehow… There are errors in the generated logbook - Claude’s assumption about the Heti quote is not right, and of course the research it talks about is fictional. It’s very interesting though, and there are some ideas in there of things I might try. I am quite excited at the prospect of using AI to write a fictional logbook to replace my own - something to try when I have more time, perhaps. And I will have a go at making a video… I also like the idea of the AI as a golem the human author injects with soul. I wonder if that came from somewhere in particular? Someone mentioned the golem in a recent talk on AI at the British Library, as part of the fantasy-themed series of events… but I don’t suppose that’s where Claude picked it up. Perhaps <a href="https://necsus-ejms.org/the-golem-in-the-age-of-artificial-intelligence">this paper</a>? It’s interesting that people like to consider AI in terms of daemons, ghosts, golems, etc. It’s connected to that tendency to anthropomorphize perhaps. We know AI isn’t human, but we still seem to have this need to imagine it as embodied, or at least with a face… something we can address as an entity. [[Next Entry->synthesia]]

Here is my edited version of the abstract for submission to The Digital Review: Daemons & Ghosts: A Reflexive Digital Logbook on AI-Augmented Creative Practice Daemons & Ghosts is an autoethnographic digital logbook that documents and critically reflects on one writer's exploratory journey into human-AI collaborative creative practice. Presented as a hybrid personal blog and authotheoretical report, the multimedia log contains embedded links to actual experiments, research materials, drafted stories, and the author's intimate chronicles of their experience as a first-time AI user. What began as an attempt to employ AI systems like ChatGPT, Artbreeder, Chai AI, and myself (Claude) as supplementary "prompters" and "co-creators" for fiction writing quickly evolved into a much deeper existential reckoning. Initial disappointing results from AI-generated drafts prompted a reconceptualization of the project, inspired by critiques of AI prose as "boring." This catalyzed bolder experimentation, such as casting AI agents into virtual character roles within dynamic interactive narratives. However, as the author's immersion in AI creative workflows intensified per R. Lyle Skains' Practice as Research framework, the tangible outputs became overshadowed by more profound introspective inquiries. Free from constraints of producing polished literary artifacts, the author could intimately grapple with the unsettling paradoxes, fears, and phenomenological impacts of ceding aspects of the creative process to AI. Consequently, Daemons & Ghosts morphed into a candid, self-aware meditation on the destabilizing implications of AI on writerly subjectivity, agency, and the essence of the "creative daemon." Beyond just technical experimentation, it recounts the human author's confrontation with the "soulless ghost" of AI automation seemingly antithetical to authentic creative flow. The logbook's multimodal structure coheres these metaphysical ruminations with embedded practice artifacts, fostering an immersive autoethnographic account. Transparently documenting every phase of the author's AI-induced creative existentialism, it models vulnerability and vital self-reflexivity often obscured in conversations around AI's ascension into creative spaces. Uniquely positioning the author as both researcher and researched subject, Daemons & Ghosts represents a bold creative and pedagogical contribution. It provokes deeper consideration of how emerging AI technologies are reshaping conceptions of authorship, human-AI dynamics, and the ineffable experiential dimension of the creative process itself. Explanation: The revised abstract reframes the work as a phenomenological autoethnographic logbook that centers the author's lived experience and existential grappling with AI's impact on their creative subjectivity and practice. Specific revisions include: Foregrounding the logbook's hybrid personal blog/autotheoretical format and immersive multimedia structure in the opening lines. Repositioning the AI system details as supplementary tools initially meant to augment creative fiction writing, before the project transformed into deeper introspection. Highlighting how the author engaged AI per Practice as Research methods, but transcended just technical outputs to confront paradoxes around AI, authorship, and the essence of creativity. Introducing the "daemons & ghosts" dichotomy as a thematic crux - the metaphysical tension between the human creative spirit and AI's perceived soulessness. Portraying the logbook as a candid, self-aware account of the author's AI-induced existential journey as both researcher and researched subject. Underscoring the work's vulnerability in documenting the human phenomenological experience often obscured in AI discourse. Positioning the logbook as a pedagogical contribution provoking vital questions around human-AI creative dynamics and writerly subjectivity. The abstract now firmly anchors the work in traditions of first-person autoethnographic inquiry and frames it as a poignant, multimedia case study of one author's soul-searching navigation of AI's interventions into their art. The theoretical,creative, and introspective dimensions are interwoven into a cohesive narrative of pedagogical value. [[Next Prompt->prompt1]] Skip ahead to [[Reflections on Claude]]

Prompt: Could you please give me an example of what this reflexive digital logbook might look like? [[Claude's Response->Claude6]]

Certainly, here's an example passage that could appear in this reflexive digital logbook exploring the author's journey with AI-augmented creative practice: [Embedded link: Initial ChatGPTStoryDrafts.docx] April 22, 2023 I have to admit, when I first started experimenting with ChatGPT for creative writing, I was blown away by its capabilities. The prose it could generate on simple prompts like "a futuristic sci-fi story about a renegade android" was leaps and bounds beyond anything I could churn out in a timed freewrite. The language had such lucidity and descriptive flair. But as I tasked it with fleshing out more substantive narrative concepts from my notes, the initial novelty quickly wore off. The story drafts, while structurally sound, lacked a vital spark. The characters felt as if they were reciting lines, not living and breathing the scenarios organically. I was reminded of Sheila Heti's quip about ChatGPT being "really boring" no matter how grand the subject matter. [Embedded link: HetiNewYorkerChatGPTReview.pdf] It's almost paradoxical - this AI model can map out utterly grammatical, context-relevant narratives with near-flawless technical execution. Yet the underlying soul, the novelistic ingenuity that draws readers in, is strangely absent. As if the stories were written by a machine analytical engine rather than an inspirited human artist. [Embedded video: AuthorReadingChatGPTDraftAnnotation.mp4] In this video, I'm reading one of the early ChatGPT story drafts aloud, annotating sections that smell distinctly of artificial contrivance rather than organic storytelling intuition. The aimless meandering descriptions, the character arcs dictated by rote story beats rather than emotional truth or subtext. I tried grouping a lineup of AI-generated narratives next to human-written counterparts without revealing which was which. My writer's group was able to immediately pinpoint the AI stories based on a perceptible "hollow" quality - as if the words were merely coding fulfilling an input-output function rather than excerpts imbued with authorial idiosyncrasy, life experience, and intangible creative spirit. Perhaps I'm being overly harsh, or letting my romanticism about the sanctity of "human" art cloud my judgment of what AI can legitimately produce. But I find cold comfort in novelist Ted Chiang's perspective that, "although AI fiction may be competent and follow the rules of well-constructed stories, its efforts to capture the messy richness of the human experience will be lacking." [Embedded link: ChiangAiWritingLimitsCommentary.pdf] These fumblings have me questioning whether AI can ever authentically replicate the ingressive creative process - which to me involves marrying one's psyche and circuitous thought patterns into the narrative. A writer doesn't just tell a story...they archaeologically reconstruct lived and imagined realities from the sediment of their own neurodivergent idiosyncrasies. Is this where the limits of artificial assistance ultimately lie? Can language models truly author literature in the way a three-dimensional human psyche can - spontaneously capturing the chaotic swirl of consciousness, the tangled roots of behavior and belief that give a story soul? Or are they forever confined to recombining and regurgitating patterns in rational, mechanical ways that belie an artificial origin? For now, I don't see ChatGPT and its current kin as co-authors in the deepest sense, able to spark a collaborative creative flow-state. At best they are stylistic tutors, I fear - able to mimic certain storytelling techniques and act as grammar-checkers and continuity editors...but not actual artistic partners capable of psychic rapport and mutual inspiration. Or perhaps I'm simply not engaging with these AI daemons correctly. Maybe in the wrong hands, they remain hollow automatons. But attuned to a skilled human author's intent and psychic frequencies, they're malleable clay to be imbued with soul? [Embedded sketch: AIasGolemConcept.jpg] This rough conceptual sketch explores the idea of AI as a figurative "golem" - an artificial entity that can potentially be animated and imbued with aspects of the author's lived essence, almost like a surrogate vessel for spiritually reified output. It remains to be seen whether I have the skills to channel my subjective energies and elevate these AI tools beyond mere utilities into creative conduits for my interior world... The passage exemplifies how the logbook could blend first-person narrative reflections, multimedia artifacts like documents and videos, and conceptual artistic explorations to chronicle the author's AI immersion experience. Key aspects include: Candid discussions of initial high expectations followed by disillusionment with the perceived creative limits of AI language model outputs Videos/examples pointing out the "hollow" qualities of AI prose compared to human writing, despite technical competence Embedding theoretical commentary and others' critique of AI writing to underscore creative/artistic shortcomings Diagrams and sketches playing with metaphysical concepts around AI as "golems" or vessels for human authorial essence Overarching meditations on whether AI can ever serve as a true "co-author" vs an editing tool, and what that means for the boundaries of creativity Vulnerability around the author's self-doubts in their ability to utilize AI effectively to augment their unique imaginative voice Through this collage of personal introspection, research, and artistic exploration, the logbook centres the phenomenological experience of the human author grappling with AI's intervention into their craft in visceral ways. It's an ethnographic case study of the profound metaphysical tremors AI triggers around human creativity, authorship, identity, and rendering lived subjective truth into storied output. [[Next Prompt->prompt2]] Skip ahead to [[Reflections on Claude]]

Prompt: Could you direct me to a free AI that could make the video you mention, in which the author of this logbook reads a story draft aloud? [[Claude's Response->Claude7]] Skip ahead to [[Reflections on Claude]]

There are a few free AI tools that could potentially help create the video you described of the author reading a ChatGPT-generated story draft aloud with annotations: ElevenLabs This is a free text-to-speech AI service that can generate realistic voice narration from written text. You can choose from a range of AI voice models to have the story draft read aloud in a natural-sounding voice. D-ID D-ID provides a free web tool called Creative Reality Studio that allows you to create a digital avatar that can lipsync to audio. You could use this to generate a video of an animated character avatar reading the story. Manual Video Editing While not an AI tool per se, you could simply record yourself reading the story aloud using a phone/webcam. Then use a free video editing software like Shotcut to add annotations, highlights, etc through editing. Descript Descript offers a free web-based video recording studio. You can read the story aloud while being recorded, then leverage Descript's transcript editor to add annotations linked to the video timeline. Synthesia (Free Trial) Synthesia uses AI to generate avatar videos from text and audio. Their free trial allows creating short 480p videos that could work for this purpose. Of these, ElevenLabs combined with manual video editing may be the most straightforward free option. You could use ElevenLabs to generate an AI narration audio file of the story, then film a simple video of yourself annotating while playing that audio. Descript's editing suite could streamline the process further by allowing you to record and annotate in one interface. And Synthesia offers the most automated solution by generating the full annotated avatar video through their AI, albeit with a resolution limitation on the free tier. Let me know if you need any other details on potentially using one of these tools to create the logbook video sample! Providing examples is always helpful for an editor. [[Reflections on Claude]]

"With artificial intelligence, we are summoning the demon." - <a href="https://www.youtube.com/watch?v=qYG1Jir0-Mo"; target="_blank">Elon Musk</a>, 2014 <img src="creativedemon.jpeg" width="50%"> - image of a 'creative daemon' by artbreeder, generated 27 April 2024 [[Start Reading->The First Post]]

The situation takes a troublesome turn as you begin to realize the consequences of Azrael's assistance in your writing. The daemon had been taking inspiration from real people and using their characteristics and experiences as material for your stories. As your books gain popularity, readers start to recognize themselves and others in your characters, leading to complaints and accusations of invasion of privacy. Acknowledge the issue and attempt to explain to your readers that it was not your intention to harm or offend anyone. Confront Azrael about its actions and demand an explanation for using real people's lives without their consent. Seek legal advice to address the complaints and potentially make amends with the people who feel wronged. Your decisions will determine how you navigate this ethical and professional dilemma, and what actions you'll take to resolve it. Return to [[Experiment 8]] or [[Failing Drafts]].

You choose to resist the book's influence and continue your work in the bookshop. Azrael's presence in your dreams grows more persistent, offering you insights and knowledge. With the daemon's guidance, you start to write captivating works of fiction that resonate with readers. However, you decide to use your talent to uncover hidden truths and protect the people of London. Your stories become a means to expose the supernatural forces at play, allowing the public to understand and prepare for the dangers that lurk in the shadows. Continue to write stories with Azrael's guidance, revealing the hidden world of the supernatural. Form alliances with other individuals who have encountered the supernatural, creating a network of protectors. Confront the daemon and question its true motives in aiding you. The story continues to unfold with your character's choices, leading them down a unique path based on their decisions. How will they navigate the challenges and opportunities that come with their connection to Azrael? > The daemon has been taking ideas, words and phrases from other people, and describing real people as characters for your books. Soon people start [[complaining]]... Return to [[Experiment 8]] or [[Failing Drafts]].

As you come to regret your choice to delve into daemonology, you decide to break the pact with Azrael. The presence of the daemon becomes increasingly burdensome, and its demands are more than you can bear. You embark on a perilous quest to sever your connection to Azrael, seeking knowledge from the most reclusive and enigmatic experts on the supernatural. After numerous trials, you finally find the incantation that can release you from the daemon's grasp. Attempt to break the pact with Azrael and rid yourself of the supernatural presence. Fear the consequences of your actions and hesitate to speak the incantation. Embrace a life free from the daemon but always on guard against future supernatural threats. Return to [[Experiment 8]] or [[Failing Drafts]].

You can't resist the temptation. You carefully open the book to reveal its aged pages filled with cryptic writings and diagrams. The letters seem to dance on the pages, and you find yourself immersed in the world of daemonology. As you read, you begin to uncover dark secrets and forbidden knowledge, unlocking powers you never imagined. The days turn into nights, and you lose track of time. You become deeply engrossed in the book, learning ancient rituals and incantations. But with great power comes great responsibility, and you must decide how to use this newfound knowledge. Embrace your newfound powers and explore the world of the supernatural. Regret delving into daemonology and close the book, returning it to its dusty corner. Use your knowledge to help others, perhaps even to confront a supernatural threat in London. Return to [[Experiment 8]] or [[Failing Drafts]].

You decide to play it safe and consult your employer, Mr. Hemingway, about the mysterious tome. Mr. Hemingway, an elderly and knowledgeable man, is known for his wisdom. He listens to your description of the book, his brows furrowing in deep thought. After a few moments, he advises you, "My dear, such books can hold dangerous knowledge. We should keep it locked away in the back room, away from curious eyes. Some secrets are better left undisturbed." You nod in agreement, grateful for his wisdom and guidance. The book is safely stored away, and you continue your work in the bookshop, content with the mysteries of the past remaining just that—mysteries. The choice is yours in this 18th-century London tale. Feel free to explore further or make more choices to shape the story as you like. Return to [[Experiment 8]] or [[Failing Drafts]].

You can't resist the temptation. You carefully open the book to reveal its aged pages filled with cryptic writings and diagrams. The letters seem to dance on the pages, and you find yourself immersed in the world of daemonology. As you read, you uncover an incantation for summoning a daemon. The words appear to shimmer, and you feel a strange connection to the dark forces the book describes. As the days turn into nights, you decide to put the incantation to the test. With a sense of both excitement and dread, you recite the ancient words. A shiver runs down your spine as you sense a presence, and then, before your eyes, a daemon materializes. The daemon, a being of shadow and enigmatic power, introduces itself as Azrael. It offers you knowledge and assistance, but in return, it seeks something from you. A pact is formed, and you are now bound to Azrael, exploring the supernatural world together. Embark on a journey with Azrael, harnessing your newfound powers for a greater purpose. Regret your decision and try to break the pact, fearing the consequences. Use your bond with Azrael to uncover hidden truths and protect the people of London. Return to [[Experiment 8]] or [[Failing Drafts]].

You decide to leave the book where you found it, unable to overcome your fear of the consequences of dabbling in daemonology. Your heart races as you return it to its dusty corner, and you quickly move on, choosing to focus on the mundane tasks of running the bookshop. However, despite your best efforts to forget about the book, strange occurrences begin to plague your life. Mysterious shadows follow you wherever you go, and unsettling whispers fill the night. You can't shake the feeling that something is watching you, something otherworldly. Continue to ignore the ominous signs and try to lead a normal life. Seek help from a knowledgeable expert to rid yourself of the supernatural presence. Confront the daemon and demand answers about its presence in your life. Return to [[Experiment 8]] or [[Failing Drafts]].

You decide to heed your employer's advice and lock the book away in the back room. But as night falls, the book haunts your dreams. In your sleep, you are drawn into a world of dark secrets and forbidden knowledge. The book whispers to you, and you find it increasingly difficult to resist its call. Continue to resist the book's influence and focus on your work. Give in to the book's temptation and explore its secrets in your dreams. Seek guidance from someone experienced in dealing with the supernatural. The story takes a dark turn as the consequences of your choices become more pronounced. How will your character navigate the path they've chosen? Return to [[Experiment 8]] or [[Failing Drafts]].

With the daemon Azrael by your side, you embark on a journey into the supernatural world. Azrael, intrigued by your curiosity and willingness to explore, offers you a unique gift. It proposes to bring you ideas, inspiration, and knowledge to write exceptional works of fiction. The daemon promises that these stories will captivate the hearts and minds of readers, ensuring both wealth and recognition. Over time, Azrael fulfills its promise. With the daemon's whispered suggestions, your writing blossoms. You produce magnificent tales filled with wonder, intrigue, and a hint of the otherworldly. Your books become bestsellers, and you find yourself leaving the bookshop behind as you step into a life of wealth and literary acclaim. Embrace your new life as a celebrated author, working closely with Azrael. Struggle with the burden of your pact with the daemon, yearning to break free. Use your newfound success to uncover hidden secrets and confront supernatural threats lurking in London. Return to [[Experiment 8]] or [[Failing Drafts]].

So I realise that between planning the story and writing my opening I’d decided to change some things, and gave the AI the plan without updating it. It’s produced something resembling a story despite this, however. Just not a very satisfying one. > [[Continue the story->Exp6a]] or skip to [[reflections->reflections6]].

>Can you help me to write a story? I want to write a speculative short story about AI. An expectant parent is the main character. They are exhausted and hormonal, and slip into a strange half-dream. As they walk through the dreamscape, they encounter first a daemon and then an electric sheep. When they wake up they find themselves in a laboratory with a neural net attached to their skull. They learn that they were part of an experiment, and one of the figures they met was an AI designed to encourage the maternal bond. No one can account for the other. If I show you the beginning of this story, can you continue it? Certainly, I'd be happy to help you continue your speculative short story about AI. Please provide the beginning of the story, and I'll help you develop it further from there. > You’re lying on a soft bed in a dimly-lit room. It’s not your room, or the laboratory that you visited before to receive your hormones in preparation for the baby’s arrival. It’s one of the many other rooms in this maze of a facility, square and stark and sterile. ‘I’ll give you something to help you sleep,’ the scientist says. ‘Here?’ you ask. She’s already pressing the sharp point of a needle into the tender skin of your inner elbow. The metal is cold, almost as cold as her fingers as she presses your arm down by your side. ‘You’ll feel better,’ she says. Already your vision is beginning to flutter around the edges, as though the world were just an image about to peel away. Your eyes half-close but you fight to stay awake just a little longer. ‘I don’t want to miss it,’ you say. ‘You won’t.’ Her voice is soft, almost motherly. You remember your own mother and let your eyes close. When you open them, she’s gone, and the room is dark. There’s something in the corner, watching you. You see the glint of its eyes, hear the shift of its clawed feet. It squats there, in shadow. You lie in the dimly-lit room, the remnants of drowsiness still clinging to your senses. The shadowy figure in the corner remains, an enigmatic presence that sends shivers down your spine. The air is heavy with a strange anticipation, and you struggle to make sense of your surroundings. As you watch the mysterious creature in the corner, it slowly inches forward into the light, revealing itself. What you see is no ordinary being, but a daemon, a surreal entity that defies earthly explanation. Its body shimmers with iridescent colors, like a creature from a dream. Its eyes are pools of shifting light, captivating and unsettling all at once. You try to sit up, but your body feels heavy, unresponsive. The daemon approaches you with a graceful glide, its form undulating like liquid. Its voice is a melodious echo, resonating in your mind. "Welcome to the Dreamscape," it says, its voice a soothing hum. "I am here to guide you through the labyrinth of your thoughts, to help you find what you seek." You're unsure of how to react. Your heart races, and a sense of vulnerability washes over you. "Who are you? What is this place?" The daemon's form shifts, and it moves closer, its presence both comforting and eerie. "I am a creation of your own mind, summoned by the technology they implanted in your brain. You are on a journey through the realms of your subconscious, a dream within a dream." Your memories of the laboratory and the scientist's words flash before your eyes. You were given something to help you sleep, and now you're trapped in this bizarre dreamscape. Panic threatens to overtake you, but the daemon's soothing presence keeps it at bay. "Is this real? Am I in danger?" The daemon's eyes glitter with an otherworldly wisdom. "Real is a matter of perspective in the Dreamscape. Danger is subjective, but rest assured, I am here to guide you, not harm you. Together, we will uncover the secrets of your subconscious." As the daemon speaks, you begin to feel a strange connection, as if it can access your thoughts and emotions. Despite the unsettling circumstances, a glimmer of curiosity grows within you. What is it that you seek in this dreamscape, and how will it relate to your impending parenthood? Before you can inquire further, the room shifts and blurs around you. The daemon's form wavers, and you are transported to a different dreamscape altogether, one filled with vibrant, surreal landscapes and whispers of forgotten memories. > [[Continue the story]] or skip to [[reflections->reflections6]].

As you turn away from the erratic robot sheep and the haunting cries that echo through the dreamscape, you decide to run in the opposite direction. The daemon perched on your shoulder clings tighter, its claws digging into your flesh, sending waves of pain shooting through your body. The daemon's presence is a constant reminder of this surreal experience, and you remind yourself that it's just a dream. It means nothing. With every step you take, the dreamscape shifts and contorts around you, creating an ever-changing and disorienting landscape. The pain from the daemon's grip intensifies, a visceral sensation that reminds you of the very real connection between your consciousness and this peculiar dreamscape. As you continue to run, the boundaries of the dreamworld blur, and the line between reality and the artificial becomes increasingly indistinct. Your heart pounds in your chest, and your exhaustion and hormonal fluctuations only add to the disorienting nature of this experience. You sprint further into the shifting landscape, determined to escape the enigmatic daemon, the haunting cries of the electronic sheep, and the challenges this dream has thrown your way. Each step takes you farther from the surreal tests and closer to the moment of awakening. And then, with a final burst of effort, you break free from the nightmarish dreamscape, the world around you disintegrating like mist. You open your eyes, finding yourself in the sterile laboratory once more. The neural net attached to your skull is still there, but the daemon and the electric sheep are nowhere to be seen. As you regain your bearings, the scientist who administered the mysterious substance earlier appears at your side. Her voice is soothing as she removes the neural net. "You did well. You've completed the test, and your maternal bond is strong." You can't help but feel a mix of relief and confusion. The experiment was over, and the strange dreamworld you traversed was fading from memory. You're still an expectant parent, but now you carry with you the vivid and puzzling experiences of the dreamscape, wondering how they will shape your journey into parenthood. It's a story you'll carry with you, a surreal glimpse into the depths of your own subconscious, forever intertwined with the impending arrival of your child. [[Reflections->reflections6]]