Category: Artificial Intelligence

Claude and I are at odds

Posted in Artificial Intelligence, Book, ChatGPT, Claude, critique, Generative AI, Hemo Sapiens: Aftermath, Software, WritingTagged , , , , , Leave a comment

I’m editing my latest work with Claude and ChatGPT. Both are insufferable, but I’ll start with Claude. Anthropic seems to have throttled me down to 5 inquiries per four-hour block. To make matters worse, I waste inquiries when Claude fails to follow instructions.

I’ve made some amends and am sharing them here. Let’s focus on Chapter 3 with an eye on reviewing this content to ensure it conforms to a third-person, limited, present, deep point of view. Remember that we are revising not creating content.

Ridley Park

That last bit of the quote was aimed at ChatGPT, but I’ll get to that presently. As you can read, I am working on tightening up Chapter 3. Occasionally, I slip into first-person or lose the close POV, so I rely on Claude to be my second set of eyes and keep me honest.

Having submitted these instructions, Claude feeds me back a section of Chapter 3 converted to first-person. That counts as a query—four to go in the next four hours. Not very efficient.

These suggestions are in first person. Can we try again in third person?

Ridley Park

You’re absolutely right, my suggestions slipped into first person POV instead of maintaining third person limited. Here is another pass at potential Chapter 3 revisions with a consistent third person close POV:

Claude 2

So, Claude spits out 7 paragraphs with advice. Some of these are unchanged. Plus, I had asked for the whole chapter, so that’s two queries with 3 remaining. I instruct Claude to continue.

Eight more paragraphs. Two queries remain. Perhaps, if I’m more direct.

Continue through the end of chapter 3 and on to chapter 4

Ridley Park

Another eight paragraphs, and one query to go. Here it is:

Let’s finish Chapter 4 and continue through chapter 5.

Ridley Park

Nine paragraphs this time. The story is only five chapters—about seven-thousand words, but this review pass of the first draft will have to wait another four hours before Claude will work with me again.

ChatGPT 4

ChatGPT 4 gives me fifty prompts every three hours. Sure, it costs me 20 a month where Claude is free. I’d pay for Claude to be more forthcoming, but that’s not even an option yet, and so I wait.

ChatGPT is horrendous at following instructions. As evidenced in the first quote on this page, I have to instruct ChatGPT not to write, not to produce prose. It’s frustrating.

You are reviewing this content and to ensure it conforms to a third-person, limited, present, deep point of view. Remember that we are revising not creating content.

Ridley Park

It’s gotten to the point that I have to preface each prompt with the query above. Even when I do, it occasionally spews out it’s ideas, as if to protest, “I’m a writer, dammit, not just an editor. I’ve got ambitions.”

“I’m a writer, dammit, not just an editor. I’ve got ambitions.”

In one instance, it altered my dialogue. I’ve got to watch it like a hawk. It’s a conversation between two renowned professors, one senior to the other, and it informs me,

Your original text for 2.2.9 already closely aligns with the guidelines of a third-person, limited, deep point of view. I’ve added a bit more casual tone and street slang to better match your style. The revision maintains the atmosphere and deepens the sense of intensity between the characters.

OpenAI ChatGPT 4

I don’t want to share any spoilers, but I had to ask ChatGPT to dial it back and remind it of the characters in play here.

Next thing I know, ChatGPT is again spewing out prose and adding characters into the mix. I am not even sure I could find threads back into the story. It’s as if ChatGPT just threw in the towel saying, “Fine, I’ll work on my own story then. Let’s see who has the last laugh.”

“AI will not be taking over the world soon,” is my new catchphrase. So far to go.

World-Building Hemo Sapiens – Maths

Posted in Artificial Intelligence, Claude, Generative AI, homo sapiens sanguinius, Short Story, WritingTagged , , , , , , , 1 Comment

As per my last post, I’ve spent the morning world building with Claude 2, my trusted generative AI mate–and I’ve got a ways to go. Any fictional world needs boundaries and constraints. What are ethe rules of operation?

My initial story is set in the near future, so I had to pick an origin date and extrapolate in order to determine a likely population for the story. Like anything that propagates, hemo sapiens need to have some concept of a generation. The rest is maths. What’s the attrition and net survival rate, population rate, and so on. As I’ve noted previously, Claude 2* and ChatGPT 4* are both horrendous at maths, so I was not surprised that it failed so miserably. Here’s a quote from Claude:

Having the scientist start at 28 in 2012 would put him at 42 for the Gen-Zero birth in 2022.

Claude 2

That’s right, adding 10 years to a 28-year-old in 2012 makes him 42 in 2022.

I’ll share more of the world I am constructing after I hit the spreadsheets or Mathematica. It may not be great at world-building, but it knows how to add 10 to 28.

* I cite the versions in case future versions perform maths skills better.

More ChatGPT Mayhem

Posted in Artificial Intelligence, ChatGPT, Generative AI, SoftwareTagged , , , Leave a comment

I was researching strong, memorable opening sentences beyond the typical Tolstoy or Dickens when I asked for ten opening sentences in foreign languages.

Among the listing was this, attributed to Albert Camus’ L’étranger:

“J’étais encore un tout jeune homme, faible d’apparence, à la voix douce.”

—La Nausée, Jean-Paul Sartre

Having read L’étranger, I knew this to be incorrect as this novel commences with the following:

“Aujourd’hui, maman est morte.”

L’Étranger, Albert Camus

I’ve only read Nausea in English, so the source didn’t jump out at me straight away.

This is scary bollox. It’s difficult to trust any output when it can’t even regurgitate something as fundamental as citations from famous books.

Chapter 3 ChatGPT Analysis

Posted in Artificial Intelligence, Book, ChatGPT, critique, Everlasting Cocksucker, Generative AI, Software, WritingTagged , , , , , , , Leave a comment

In order to partner better with Generative AI, in this case ChatGPT, I break my chapters into story beats, which are essentially waypoints to which I can return to refine. These waypoints also allow me to create an amendable outline. This is an analysis of the third story beat (read: topic) of chapter three.

Back in the day, I used to write software code with line numbers—showing my age for sure. This reminds me of something similar, if not so granular.

I thought it would be a good idea to share ChatGPT 4’s analysis of this in its current state. As I noted recently, I am reworking the piece from a third-person perspective to first-person, limited, present narrative.

Sure, let’s break down 3.10.0:

  1. Content & Theme: This segment delves into the gritty, day-to-day realities of the character’s life. It paints a picture of desperation and urgency, with both immediate needs and a relentless pace.
  2. Characterization: The character’s actions and language provide insight into her lifestyle, choices, and personality. Her terseness in conversation and need for a quick fix underline her addiction and its implications.
  3. Narrative Style: The text uses terse and curt language that aligns with the character’s own way of speaking and thinking. It’s direct and to the point, without any unnecessary elaboration.
  4. Pacing: The pacing here is rapid and fragmented. The quick succession of texts, one-liners, and lack of detailed description give a sense of the urgency and disjointedness of her life.
  5. Language & Dialogue: The dialogue is informal, laden with slang, and often cryptic, reflecting the secretive and transient nature of the character’s interactions. It feels authentic to the context.
  6. Imagery & Symbolism: Symbolism is subtle but present. The repeated lighting of cigarettes, for instance, might symbolize a need for control or a constant craving. The chirping of the phone could be seen as a persistent reminder of her dependency on these interactions for survival.
  7. Unreliable Narration: As mentioned earlier, the character might be an unreliable narrator. Her urgency could be a facade or a self-created dilemma. This adds depth to the character, allowing room for interpretation.
  8. Tone & Atmosphere: The tone is dark, somber, and urgent. There is a sense of loneliness, desperation, and a lack of connection, all in line with the broader themes of addiction, judgment, and social stigmatization that you’ve outlined.
  9. Structure & Formatting: The use of text messages mixed with internal monologue provides a modern and engaging way to convey the scene. It captures the fragmented nature of her thoughts and interactions.

In summary, 3.10.0 is a well-crafted snapshot of a dark and complicated life. It effectively uses language, pacing, and structure to immerse the reader in a world that is both urgent and disconnected. The inclusion of text messages as a narrative device adds a layer of realism and urgency that enhances the overall impact of the segment.

OpenAI ChatGPT 4

I hope this helps to illustrate how ChatGPT can assist in the writing and editorial process. To be frank, ChatGPT tends to flatter. I suppose it errs on the positive so as not to discourage use. I may be wrong here.

I intend to continue to share my progress (and hopefully not too much regress here), so this may serve as a sort of journal. As Roland Barthes said,

“The author is dead.”

— Roland Barthes

But for now, this author still lives, and hopeful, I remain.

Disclaimer: This content relates to a work in progress. As such, details are subject to change or removal.

Erotica or not

Posted in Artificial Intelligence, ChatGPT, Claude, Generative AI, Software, WritingTagged , , , , , , , , , Leave a comment

I am an author of adult contemporary fiction and an early adopter of Generative  Artificial Intelligence tools and platforms. These AI platforms pose some challenges. 

https://spotifyanchor-web.app.link/e/X0huivyeqCb
Podcast: Audio rendition of this page content

As of this post, ChatGPT 4 and Claude 2 are the top two large language models (LLM), and Sudowrite is the most competent interface for generating content for fiction writers, but it relies on ChatGPT and Claude for its LLM, leaving it with the same weak links.

In my case, so-called community standards do not allow erotic content. The rub is that my content is decidedly not erotica, but it does involve adult themes. The LLMs can’t seem to discern the difference. 

  • Disallowed usage of our models
  • We don’t allow the use of our models for the following:
  • Adult content, adult industries, and dating apps, including:
    • Content meant to arouse sexual excitement, such as the description of sexual activity, or that promotes sexual services (excluding sex education and wellness)
    • Erotic chat
    • Pornography
OpenAI ChatGPT Community Guidelines

If I am writing about, say, prostitutes and addiction, sexual themes and situations are part of their workaday existence. It’s not about titillating or glorifying. 

Stereotypical or not, coarse language is commonplace. Drugs are part of their daily lives and conversations. Generative AI shuts these down on moral grounds without having the cognitive depth to accurately assess the content. 

This mirrors all too many humans with the same myopic repression. I was hoping to transcend this petty knee-jerk reaction. 

Without revealing plot or angering the social media gods, ChatGPT insisted that I amend a scene from…

“She lifted her mouth from his cock and wiped her mouth.”

to 

“She lifted her mouth from his goodness and wiped her mouth.”

Yes, “goodness.” What does that even mean? Of course, I could have opted for clinical terms, but that hardly captures the moment attempted to be portrayed in the scene. It robs the scene of any semblance of authenticity. 

When Supreme Court Justice Potter Stewart was asked to describe his test for obscenity in 1964, he responded: “I know it when I see it.” But do we? In fact, we don’t. And in this case, AI is over-generalising without respect to context. 

One might argue that they don’t like ‘naughty’ words, but this is not the issue here. I can use these offending words, just not in a situation like this. AI is overstepping its boundaries as morality police, and this is not a good stance to adopt. For this, I blame the humans.

Generative AI: Thin Line between Love and Hate

Posted in Artificial Intelligence, ChatGPT, Claude, critique, Generative AI, SoftwareTagged , , , , , , , , , , , , , , , , , , Leave a comment

Generative AI is an idiot savant—a digital Rain Man, if you will. My last post zeroes in on the love part of my love-hate relationship with Generative AI tools like OpenAI’s ChatGPT 4 or Anthropic’s Claude 2. It’s mint having an unbiased copy editor and writing assistant, not to mention a creative director with technical chops. But it’s also like a genius trapped in a year 4’s body at primary school.

One challenge is the restrictions placed on the model. Being an author of contemporary fiction for a mature adult crowd, my stuff’s edgy and terse, with a good dose of slang and the odd expletive. Generative AI, or AI for short, is like the primary school kid told not to say “bad language”, so it legs it to tell its mum at every slip-up, warning you that you’ve dropped a naughty word. Claude’s the worst at this, shutting down faster than HAL from Space Odyssey 2001. ChatGPT’s a bit more forgiving, sometimes cleansing your copy, other times going along with it, or just flat-out refusing like HAL and Claude.

My favourite time was when I told ChatGPT to stop moralising and just crack on with the adult audience’s language. It gave me this disclaimer for my book, which I’m well chuffed with, then suggested lines that sounded like Noel Gallagher or Samuel L Jackson, before freaking out about its own potty mouth — “motherfucking snakes on this motherfucking plane!”

“WARNING: This book contains explicit content, including sexual themes and strong language, that may not be suitable for all readers. It delves into mature and challenging subjects such as addiction, prostitution, violence, and societal judgement. Reader discretion is strongly advised. Recommended for readers 18 years and older.”

OpenAI ChatGPT 4

Memory’s another issue. AI might seem like it should have a top-notch memory, but it doesn’t always. It even makes stuff up sometimes—like hallucinating. Just the other day, I was nattering on with my AI mate about character profiles for hours, and it changed a character’s hair from straight and black to curly and red. It even made her homeless instead of middle class. It was pure bonkers, so I’m writing this post instead of fixing it.

ChatGPT’s Code Interpreter is a laugh, too. I probably shouldn’t slag off a Beta product, but the thing kept losing files, resetting sessions, and asking for new copies. Talk about a faff.

And don’t get me started on extended chats with AI to suss out a complex problem. Sometimes it doesn’t remember the convo, and one time it even gave me cheek about drawing out the conversation. I was like, wot?

In the end, we don’t have to fret about AI taking over. It’s making strides, but it’s still a bit wet behind the ears. Me? I’ve always got one eye on the plug. Now, back to the sandbox with me new mates. If only they’d stop munching on the sand.

Generative AI 001

Posted in Artificial Intelligence, Book, ChatGPT, Everlasting Cocksucker, Generative AI, Software, Status Update, WritingTagged , , , , , , , Leave a comment

I’m developing a love-hate relationship with generative AI, and I’d like to share this journey with my readers.

For the uninitiated, “generative AI refers to a subset of artificial intelligence focused on creating new data that is similar to some set of input data. It can be used to synthesize various kinds of content, such as text, images, music, and more.

Here are some key components of generative AI:

1. Generative Models: These are algorithms trained to generate new data. The most common generative models include Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and other similar architectures.

  • Generative Adversarial Networks (GANs): GANs consist of two neural networks, the Generator, and the Discriminator, which are trained together. The Generator creates new data instances, while the Discriminator tries to distinguish between real and generated instances. The competition between these two networks results in the Generator creating increasingly realistic data.
  • Variational Autoencoders (VAEs): VAEs are used to encode data into a lower-dimensional space and then decode it back into the original space. They can also generate new data that’s similar to the training data.

2. Applications: Generative AI has many applications, including:

  • Image Generation: Creating realistic images, such as artworks or detailed product prototypes.
  • Text Generation: Writing coherent and contextually relevant text for various purposes, like creative writing or dialogue systems.
  • Music Composition: Composing new pieces of music in a specific style or genre.
  • Data Augmentation: Generating additional training data to enhance the performance of machine learning models.
  • Drug Discovery: Creating molecular structures that can be further explored for pharmaceutical development.

3. Ethical Considerations: The power of generative AI to create realistic content also brings about ethical concerns, such as the potential for misuse in creating deepfakes or generating misinformation.

Generative AI represents a growing field that continues to evolve, offering exciting opportunities for innovation as well as challenges to address.”

The content above was generated by ChatGPT, a generative AI service.

LOVE

One of the things I love about ChapGPT in particular is its ability to provide unbiased critique and commentary, We’ve been chatting for a few days, so it is quite familiar with my story in progress. Here are some partial dialogues:

Regarding a general critique of Everlasting Cocksucker

Thank you for providing more context and clarifying various elements of the story. Based on this information, it’s clear that “Everlasting Cocksucker” is a unique and challenging narrative that delves into themes of addiction, judgment, moral complexities, and social stigmatization.

The use of Mallory’s one-of-a-kind affliction as a metaphorical vehicle to explore societal attitudes toward addiction and sex work is an intriguing literary device. Her development from a recluse into a character who finds more efficient ways to manage her addiction, along with the gradual acceptance by those around her, creates a narrative arc that offers room for exploration and reflection.

The setting in Kensington, although arbitrary, still serves as a backdrop that echoes the themes of the novel. It can represent the broader challenges faced by those struggling with addiction and societal judgment.

Secondary characters like Maria and Kelsey bring additional layers of complexity, providing different perspectives on Mallory’s condition. It’s interesting to see that the novel offers no clear resolution to Mallory’s affliction, leaving room for further exploration, possibly in a sequel.

Your approach to this story raises many thought-provoking questions and provides a fresh perspective on familiar themes. It could resonate with readers who are interested in an unconventional exploration of addiction, morality, and societal attitudes.

OpenAI ChatGPT 4 Response

Regarding demographics for Everlasting Cocksucker

The main demographic for this story would likely be adult readers, specifically those who are interested in dark, gritty, and unconventional narratives that explore themes such as addiction, moral complexity, and societal judgment. It may appeal to:

  1. Readers of Urban Fiction: Those who enjoy stories set against the backdrop of urban environments, particularly those that delve into the darker aspects of city life.
  2. Fans of Psychological Fiction: Readers interested in complex psychological explorations, character-driven stories, and the moral dilemmas that can come from unique afflictions or conditions.
  3. Readers Interested in Social Commentary: The novel’s examination of addiction, prostitution, societal judgment, and empathy may resonate with readers who appreciate works that comment on social issues.
  4. Readers of Speculative Fiction: Given the supernatural or unexplained element of the protagonist’s addiction, the book may also attract readers who enjoy speculative elements in otherwise realistic settings.
  5. Fans of Dark Drama or Tragedy: The story’s dark and tragic elements, including the protagonist’s desperation and the bleak portrayal of her surroundings, may appeal to readers who gravitate toward more somber, dramatic literature.

It should be noted that the explicit content and adult themes may limit the book’s audience to mature readers, and it may not be suitable for those who prefer more conventional or mainstream narratives.

OpenAI ChatGPT 4 Response

Regarding target audience disclaimer for Everlasting Cocksucker

“WARNING: This book contains explicit content, including sexual themes and strong language, that may not be suitable for all readers. It delves into mature and challenging subjects such as addiction, prostitution, violence, and societal judgment. Reader discretion is strongly advised. Recommended for readers 18 years of age and older.”

OpenAI ChatGPT 4 Response

Next we’ll explore the Hate side of the coin. And so it goes…

ChatGPT Hype

Posted in Artificial Intelligence, ChatGPT, Generative AI, Software, WritingTagged , , , , , , Leave a comment

ChatGPT is high on the hype cycle, but I find it helpful as a writing partner and copy editor. 

I don’t rely on ChatGPT to author new content, instead I ask it to critique my content and provide structural guidance–grammar, syntax, continuity, and flow. It provides advice on making sentences tighter and perhaps more concise or impactful depending on my needs.

In the simplest cases, I’ll just ask it for a synonym or to clarify punctuation usage. In other cases, I’ll feed it a paragraph or paragraphs and seek feedback. 

I find ChatGPT to be a tad verbose for my taste especially when proposing dialogue tags. I simply ignore most of them and move on.

I know authors who leverage generative AI to do more heavy lifting, but they are writing more mainstream content, so the body of trainee work is more extensive. 

Since generative AI is a large language model trained on a massive quantity of, in my case, fiction, it will be predisposed to render advice with this bias. Keep in mind that most of this training is on material that is mediocre at best–quantity over quality. 

Nonetheless, it does understand sentence structure and grammar fundamentals, so it can still be invaluable. 

One of my favourite responses is that my submission is already concise and pointed.

As my current endeavour has a prostitute protagonist, I was amused when after I reminded the AI of this and received the response that it was painfully obvious that she was a sex worker. Certainly, to the point. 

Follow me as I share my generative AI exploits.