Split from: What's the best site to copyright your work that isn't a potential scam?

[PJ Post]

How do I go about hiring beta readers for niche-specific reading? I posted something in the proofreading area, but is there a better place to ask for it? Since you're mentioning getting my ducks in a row.

Nothing can replace an honest human reaction, but before you get there...

What we all need is a pro NY editor from the 1960s, someone like Maxwell Perkins - but that isn't going to happen. The next best thing - ChatGPT.

I am not suggesting you use AI to write anything - just to provide editorial commentary that you can either listen to or ignore, just like any other editor.

Here's the prompt:

Pretend you are Maxwell Perkins, the legendary editor of Fitzgerald, Hemingway, and Wolfe. Read the following story or chapter as if I were your author. Provide an editorial review that covers not only grammar, punctuation, homonym errors, and overused words, but also character depth, thematic resonance, developments, pacing, logical consistency, and potential plot holes. Evaluate voice, tone, and style for literary strength, and note where trimming or expansion might help. Be both encouraging and exacting, with Perkins's balance of empathy and precision, pushing me to make the work its best.

And then you upload the chapter for it to review. And it's safe, nobody at OpenAI cares about our stories.

Next Prompt:

This is a marketing check to make sure your story is genre appropriate.

Evaluate my story for fit within the Christian fantasy genre. Compare its voice, language, tone, themes, and use of tension to established works such as The Chronicles of Narnia by C.S. Lewis, The Lord of the Rings by J.R.R. Tolkien, A Wrinkle in Time by Madeleine L'Engle, The Pendragon Cycle by Stephen R. Lawhead, and The Circle Series by Ted Dekker. Identify where my story aligns with genre expectations, where it diverges, and whether those divergences strengthen or weaken its position as Christian fantasy. Offer feedback on clarity, style, and overall genre appropriateness.

Insert whatever other books and authors you would prefer.

Also, you can run these prompts again based on a different editor(s) or books to stress test the AI.

___

I run all my stuff through GPT like this - just to make sure I'm not missing something stupid, like someone's eye color I forgot or a plot hole. AND THEN, after you've fixed everything - have another human read it. You won't need as many beta readers either. Oh, and I still have other humans read everything before it sees the light of day.

___

NOTE: like any editorial advice, be careful not to let them edit the soul out of your work. It's okay to reject editorial advice, be it human or AI.

___

On copyright: as noted, the work is immediately protected upon being fixed in a tangible medium of expression - and the only reason you need to register a copyright is if you plan on spending time in a courtroom - either to protect your work or defend it. But - generally speaking - nobody cares until there's a boatload of money at stake. Few self-publishers ever earn courtroom money. But if you did, you could hire the lawyers you need - QED.

[TimothyEllis]

@Matt

PJ is our resident 'drunk the AI koolaid' pusher.

My reaction to what he just posted is F*CK NO!

None of the bots are yet up to MORON level.

Why would you ask a moron to do anything with your writing?

I especially would not be asking it for advice on anything to do with writing, since they've all been fed with the absolute WORST writing that is available. Not to mention, they make sh*t up all the time.

[TimothyEllis]

OpenAI offers free accounts. Try it and see for yourself. What's the worst that can happen?

They plagiarize what you put in, and give it to someone else.

[Post-Doctorate D]

OpenAI offers free accounts. Try it and see for yourself. What's the worst that can happen?

They plagiarize what you put in, and give it to someone else.

Yesterday, I saw a post on Facebook by an author.  She took a sample of one of her books and had AI examine it.  The AI determined that there was a high probability the sample was written by AI.

The thing is that she wrote and published the book before AI was available, so there's no way she could have used AI to write it.

Her point was that AI has stolen so much of our work and regurgitated it that it can now mistake original works for something written by AI.

[Post-Doctorate D]

OpenAI offers free accounts. Try it and see for yourself. What's the worst that can happen?

They plagiarize what you put in, and give it to someone else.

Also, for those that don't think AI companies aren't doing that, check the terms of service for the tool you are using.  Generally, you are agreeing to let them use your submissions to "improve" their AI tools.

[Bill Hiatt]

Yes, copyright theoretically protects your work from the moment you write it. Proving that may be the hard part. But I suppose it's true that a publication date on Amazon or elsewhere is pretty difficult to argue with. Substack posts, of course, pose different issues, because if they are serials, it's tough to copyright before the serial is finished. However, you do have a Substack publication date for each episode.

I also save any earlier drafts, just in case. They can help pinpoint a date at which work reached a certain point.

On AI, you have to decide as an author whether you want to use a tool that's built on ethically questionable use of the intellectual property of others. You also need to consider the occasional odd lapses, some of which are worse than others.

For anyone writing nonfiction, keep in mind that AI will invent facts. It's produced false case citations for lawyers and false medical studies in a recent HHS report. Given that tendency, I don't thing I'd trust it to make literary judgments.

Even in AI-generated search results, it still occasionally flubs. Just yesterday, it gave me a factual error on Greek mythology, and not so long ago it did the same with the KDP TOS. The latter is harder to understand. Also, when I was trying to check whether West LA College had an indoor pool, it told me yes and showed me a picture--which was actually of an outdoor pool at Valley College. (WLA has no pool, indoor or outdoor.) It also told me West LA has no Saturday classes. I checked one of the linked sources, which said exactly the opposite. If it can't even get simple facts right, why are we trusting its literary judgement?

As others have pointed out, AI uses your work for further training. Don't feed the beast that may one day devour you.

As PJ demonstrated, it can produce reasonable summary responses on some issues--but only if someone else has already written on them. And usually, when I check the response against the sources, the original human writing is just as good. Occasionally, AI may successful combine two pieces into one, but otherwise, it seems better to just look at the human sources. At least on Google, AI seems mostly to draw on the two or three top search results. It's not as if it actually synthesizes everything on the whole internet. (That's probably just as well, given the high and growing energy costs.)

PJ makes a good point, though, about not letting someone (or something) edit out your individual voice. I've had more problem with tools like pre-AI Grammarly in that regard than I've ever had with a human editor, though. 

[PJ Post]

As others have pointed out, AI uses your work for further training. Don't feed the beast that may one day devour you.

As PJ demonstrated, it can produce reasonable summary responses on some issues--but only if someone else has already written on them. And usually, when I check the response against the sources, the original human writing is just as good. Occasionally, AI may successful combine two pieces into one, but otherwise, it seems better to just look at the human sources. At least on Google, AI seems mostly to draw on the two or three top search results. It's not as if it actually synthesizes everything on the whole internet.

First, OpenAI allows you to opt out of having your work used in training. Sure, they could be lying, but I don't have time for conspiracy theories. So, no beast to feed here - no Bigfoot lurking in the shrubbery.

Next, I use AI as a tool. Full stop. Not to write for me - as I've reiterated ad nauseum here - but as a tool (literary, business, research, project management, data crunching, marketing, software and hardware manual, efficiencies, web search and brainstorming and collaboration - AMS/CPC strategies are next on the list).

Links to some of my work are right down there at the bottom of this post. Feel free to read a sample and decide if you think that I know how to write or not - to see if I'm professional enough to have a valid opinion on the subject. You won't hurt my feelings.

AI literary analysis of my work is absolutely amazing. It's detailed, original to my work and pretty accurate. Does it make mistakes? Sure it does. Did my paid editors make mistakes? Sure they did. This is why you stress test it, and get various summaries and then compile a final list of possible edits, which you then check against your actual work and decide what matters and doesn't. A rototiller will take your foot off if you're not careful - learn to use the tool. And then, and only then, do you get to complain about how much it sucks.

Finally, for the  - f*ck it, I don't even know how many times - AI does not plagiarize our work. It does not store our work. It does not combine various IPs to get a new thing. That's not how it works. That's not how it has ever worked. Stop reading clickbait articles about how cameras are stealing our souls AI is stealing our work.

As for AI hallucinating, GPT5 is better, but you have to be a knowledgeable user. AI is an assistant. Would you believe everything a junior researcher claimed at work? Of course not. You stress test it, you create validation prompts and doublecheck the citations, just like any rigorous review of any academic paper. It's not Data from Star Trek, even though everyone desperately wants it to be.

AI is software. It's an absolutely amazing version of Clippy. Will AI become our hyper-intelligent Overlords? Maybe, but that day is still not today.

___

Here's an AI analysis of this post:

Here's a close editorial review of your passage, broken down by category:


Themes


AI as a tool - Emphasizing utility, not replacement for human craft.

Skepticism vs. Conspiracy - Rejecting alarmist narratives ('Bigfoot', 'cameras stealing souls').

Professional credibility - Inviting readers to judge your skill by your published work.

Fallibility of all tools - Both AI and humans make mistakes; the responsibility is on the user.

Demystification - Comparing AI to mundane tools (rototiller, Clippy) rather than sci-fi fantasies.


Logical Consistency


Generally consistent: you argue AI is a tool, not a plagiarist, and compare its flaws to those of human editors.

'AI does not plagiarize, store our work, or combine IPs' - broadly correct, but stated in absolutes. More precise phrasing would avoid reader pushback (e.g., 'does not retain user data permanently for training' instead of 'does not store our work').


Tone


Conversational, blunt, sardonic.

Humor/light sarcasm ('Bigfoot lurking,' 'amazing version of Clippy').

Rhetorical repetition (- for the - f*ck it, I don't even know how many times -).

Tone is engaging, but risks alienating skeptical readers; sarcastic phrasing can come across as dismissive rather than persuasive.


Language


Mostly clear, everyday diction.

Occasional rough edges: 'f*ck it' may limit audience reach if professionalism is desired.

Some wordiness in parentheticals; breaking into shorter sentences would improve rhythm.


Accuracy


Saying AI never plagiarizes could be softened (better: 'doesn't directly lift text from sources' or 'is designed not to reproduce copyrighted passages verbatim').

'GPT-5 is better' is a subjective claim - acceptable in personal essay but worth framing as opinion.


Recommended Edits


Clarity & precision


'OpenAI allows you to opt out of training use; it does not automatically use your data unless you give permission.'

Replace 'AI does not plagiarize our work' → 'AI isn't designed to plagiarize or store our individual work'.


Tone adjustment


'f*ck it' → could be swapped for 'at this point' or 'honestly' if you want to retain casualness without alienation.

'cameras stealing our souls' → keep (funny, accessible).


Flow


Break up long parentheses: 'literary, business, research, project management...' could be smoothed into two sentences or a list for readability.


Tightening


'AI literary analysis of my work is absolutely amazing' → consider trimming 'absolutely.' Stronger without the adverb.

'A rototiller will take your foot off if you're not careful' is great imagery; you could shorten to 'Like a rototiller, it can do damage if misused.'


Other Advice


Decide whether the goal is persuasion or venting. If persuasion, tone should be slightly less combative - especially on sensitive points like plagiarism/storage.

Consider audience knowledge: Readers unfamiliar may stumble on insider shorthand like 'AMS/CPC strategies' unless briefly explained.

Stronger close: The 'Clippy' line is memorable. Let that stand as the button without adding the 'Overlords' line, which weakens the punch.

[MattT]

How do I go about hiring beta readers for niche-specific reading? I posted something in the proofreading area, but is there a better place to ask for it? Since you're mentioning getting my ducks in a row.

Nothing can replace an honest human reaction, but before you get there...

What we all need is a pro NY editor from the 1960s, someone like Maxwell Perkins - but that isn't going to happen. The next best thing - ChatGPT.

I am not suggesting you use AI to write anything - just to provide editorial commentary that you can either listen to or ignore, just like any other editor.

Here's the prompt:

Pretend you are Maxwell Perkins, the legendary editor of Fitzgerald, Hemingway, and Wolfe. Read the following story or chapter as if I were your author. Provide an editorial review that covers not only grammar, punctuation, homonym errors, and overused words, but also character depth, thematic resonance, developments, pacing, logical consistency, and potential plot holes. Evaluate voice, tone, and style for literary strength, and note where trimming or expansion might help. Be both encouraging and exacting, with Perkins's balance of empathy and precision, pushing me to make the work its best.

And then you upload the chapter for it to review. And it's safe, nobody at OpenAI cares about our stories.

Next Prompt:

This is a marketing check to make sure your story is genre appropriate.

Evaluate my story for fit within the Christian fantasy genre. Compare its voice, language, tone, themes, and use of tension to established works such as The Chronicles of Narnia by C.S. Lewis, The Lord of the Rings by J.R.R. Tolkien, A Wrinkle in Time by Madeleine L'Engle, The Pendragon Cycle by Stephen R. Lawhead, and The Circle Series by Ted Dekker. Identify where my story aligns with genre expectations, where it diverges, and whether those divergences strengthen or weaken its position as Christian fantasy. Offer feedback on clarity, style, and overall genre appropriateness.

Insert whatever other books and authors you would prefer.

Also, you can run these prompts again based on a different editor(s) or books to stress test the AI.

___

I run all my stuff through GPT like this - just to make sure I'm not missing something stupid, like someone's eye color I forgot or a plot hole. AND THEN, after you've fixed everything - have another human read it. You won't need as many beta readers either. Oh, and I still have other humans read everything before it sees the light of day.

___

NOTE: like any editorial advice, be careful not to let them edit the soul out of your work. It's okay to reject editorial advice, be it human or AI.

___

On copyright: as noted, the work is immediately protected upon being fixed in a tangible medium of expression - and the only reason you need to register a copyright is if you plan on spending time in a courtroom - either to protect your work or defend it. But - generally speaking - nobody cares until there's a boatload of money at stake. Few self-publishers ever earn courtroom money. But if you did, you could hire the lawyers you need - QED.

I'm using ChatGPT as a try, and if I had found this thing 2 years ago, I could have organized my books in a fraction of the time. Didn't know AI could be so useful.

[MattT]

How do I go about hiring beta readers for niche-specific reading? I posted something in the proofreading area, but is there a better place to ask for it? Since you're mentioning getting my ducks in a row.

Nothing can replace an honest human reaction, but before you get there...

What we all need is a pro NY editor from the 1960s, someone like Maxwell Perkins - but that isn't going to happen. The next best thing - ChatGPT.

I am not suggesting you use AI to write anything - just to provide editorial commentary that you can either listen to or ignore, just like any other editor.

Here's the prompt:

Pretend you are Maxwell Perkins, the legendary editor of Fitzgerald, Hemingway, and Wolfe. Read the following story or chapter as if I were your author. Provide an editorial review that covers not only grammar, punctuation, homonym errors, and overused words, but also character depth, thematic resonance, developments, pacing, logical consistency, and potential plot holes. Evaluate voice, tone, and style for literary strength, and note where trimming or expansion might help. Be both encouraging and exacting, with Perkins's balance of empathy and precision, pushing me to make the work its best.

And then you upload the chapter for it to review. And it's safe, nobody at OpenAI cares about our stories.

Next Prompt:

This is a marketing check to make sure your story is genre appropriate.

Evaluate my story for fit within the Christian fantasy genre. Compare its voice, language, tone, themes, and use of tension to established works such as The Chronicles of Narnia by C.S. Lewis, The Lord of the Rings by J.R.R. Tolkien, A Wrinkle in Time by Madeleine L'Engle, The Pendragon Cycle by Stephen R. Lawhead, and The Circle Series by Ted Dekker. Identify where my story aligns with genre expectations, where it diverges, and whether those divergences strengthen or weaken its position as Christian fantasy. Offer feedback on clarity, style, and overall genre appropriateness.

Insert whatever other books and authors you would prefer.

Also, you can run these prompts again based on a different editor(s) or books to stress test the AI.

___

I run all my stuff through GPT like this - just to make sure I'm not missing something stupid, like someone's eye color I forgot or a plot hole. AND THEN, after you've fixed everything - have another human read it. You won't need as many beta readers either. Oh, and I still have other humans read everything before it sees the light of day.

___

NOTE: like any editorial advice, be careful not to let them edit the soul out of your work. It's okay to reject editorial advice, be it human or AI.

___

On copyright: as noted, the work is immediately protected upon being fixed in a tangible medium of expression - and the only reason you need to register a copyright is if you plan on spending time in a courtroom - either to protect your work or defend it. But - generally speaking - nobody cares until there's a boatload of money at stake. Few self-publishers ever earn courtroom money. But if you did, you could hire the lawyers you need - QED.

I'm using ChatGPT as a try, and if I had found this thing 2 years ago, I could have organized my books in a fraction of the time. Didn't know AI could be so useful.

I wont use it to to write, that's effectively stealing, but I can see it being useful for organization. But given Timothy's and other people's feelings towards OpenAI, I can see that its best to leave it out altogether. As someone who uses a platform like Quora to debate with people. I find it extremely irritating when someone clearly uses AI to fight for them, in lieu of using their own voice. Why trust it to write for me too?

[Bill Hiatt]

Next, I use AI as a tool. Full stop. Not to write for me - as I've reiterated ad nauseum here - but as a tool (literary, business, research, project management, data crunching, marketing, software and hardware manual, efficiencies, web search and brainstorming and collaboration - AMS/CPC strategies are next on the list).

Links to some of my work are right down there at the bottom of this post. Feel free to read a sample and decide if you think that I know how to write or not - to see if I'm professional enough to have a valid opinion on the subject. You won't hurt my feelings.

I apologize if I wasn't clear. I never intended to suggest that you use AI for writing. I don't recall anyone else doing that, either. The question before is not how you're using AI but rather how--and if--anyone should. Nothing I've said was ever intended as a criticism of you as a writer. The nature of our disagreement is philosophical, or at least, I thought it was.

Finally, for the  - f*ck it, I don't even know how many times - AI does not plagiarize our work. It does not store our work. It does not combine various IPs to get a new thing. That's not how it works. That's not how it has ever worked. Stop reading clickbait articles about how cameras are stealing our souls AI is stealing our work

Yes, but AI doesn't have to plagiarize to be unethical. We all agree that LLMs would not exist without the intellectual property upon which they were trained. (And at least according to some developers, they pretty much need access to whole content of the internet forever to maintain themselves.) We should all agree, at the very least, that AI training is radically different from anything claimed as fair use up to this point. We know that at least some AI training was done using data derived from pirated databases which any reasonable person would have known were pirated. Developers used them with complete disregard of their illicit origin. They couldn't even be bothered to go and buy copies of the books they wanted to use. And from what Lynn says, it sounds as if Anthropic is going to have to pay for that.

That said, I understand that many people see AI applications as appropriate tools. That's their choice. That's your choice. The only creators who really irk me are the ones who use AI to generate content and make only minor tweaks--in which case, it's more like AI is the creator, and they are the tool.

There are a lot of different ways to use AI. I'm more comfortable with ones involving ethically sourced AI (training material creators are compensated, which some stock image providers are now doing) and using it for situations in which a human wouldn't have been hired to do the job. As you may recall, I was using it for images for a while myself. I stopped because of negative feedback. But I'm certainly not one to say that the ethics are always crystal clear in these kinds of situations.

It's great that Open AI lets you opt out of your material being used as training. Does every AI company? Regardless, feeding the beast can also be referred to an encouraging  AI companies by using their products. If people stopped using them, perhaps developers would focus more on the altruistic applications, like curing cancer. Yeah, that's unlikely. I'm still personally not going to feed the beast. Others may do as they wish.



 

[TimothyEllis]

That said, I understand that many people see AI applications as appropriate tools.

And that's where I have the problem.

They've been trained on the absolute WORST written drek that's available.

How does any of that stand up as a decent tool for writers?

It doesn't.

If they'd been very carefully fed the very best of writing, then maybe they'd make a good tool.

But even then, a tool fed all of Shakespeare and Tolkien and contemporary writers to both, is still not useful for a writer today. The language has changed, the way we write has changed, what is important to writing has changed.

The tools would need to be very carefully fed for specific uses as separate tools to be any real use at all.

Instead we've got bots fed with 50 million books that never sold, all the fan fiction since the 80's, poetry from times gone by, all the books out of copyright that are so old they're not relevant to modern writing, and then the vast sewer of the internet and social media where people are barely literate and don't care about grammar and spelling at all.

How is any of that useful as a tool for writers?

It's not.

[PJ Post]

I'm posting this again because it seems to have been missed. Erroneous and extremely misleading assertions were posted about AI and, in response, I posted an example that clearly disproved them. It matters because we have new folks here that are missing out on a huge opportunity to help them level up because of these sorts of misleading posts. That's not fair or particularly cool.

It's fine to not want to use a tool, but quite another to demonize it.

You guys are stuck in a loop where you think AI can only write our books for us or it's a push-button cover generator - and nothing else - self-publisher-centric. AI is a collaborator and an assistant that blends seamlessly with our workflow. I'm recommending AI because I use it, I get it. I understand the value. And I know lots of other folks who use it, and they're all improving efficiency, output, accomplishments. It's making them all better at whatever their thing is. And we all use it differently: different fields, different needs.

Tim feel free to move this to a new thread. I'm posting here because this is where the discussion took place.

I posted a response to those misleading assertions and then had AI analyze that response. The AI response was original to my post, specific to my post, clear, accurate and demonstrated a deep understanding of nuance, style and audience - everything you guys said it could never do. This one post demonstrates how deep AI's understanding of language goes and scarcely hints at its potential - none of which needs to be writing our books for us.

As others have pointed out, AI uses your work for further training. Don't feed the beast that may one day devour you.

As PJ demonstrated, it can produce reasonable summary responses on some issues--but only if someone else has already written on them. And usually, when I check the response against the sources, the original human writing is just as good. Occasionally, AI may successful combine two pieces into one, but otherwise, it seems better to just look at the human sources. At least on Google, AI seems mostly to draw on the two or three top search results. It's not as if it actually synthesizes everything on the whole internet.

First, OpenAI allows you to opt out of having your work used in training. Sure, they could be lying, but I don't have time for conspiracy theories. So, no beast to feed here - no Bigfoot lurking in the shrubbery.

Next, I use AI as a tool. Full stop. Not to write for me - as I've reiterated ad nauseum here - but as a tool (literary, business, research, project management, data crunching, marketing, software and hardware manual, efficiencies, web search and brainstorming and collaboration - AMS/CPC strategies are next on the list).

Links to some of my work are right down there at the bottom of this post. Feel free to read a sample and decide if you think that I know how to write or not - to see if I'm professional enough to have a valid opinion on the subject. You won't hurt my feelings.

AI literary analysis of my work is absolutely amazing. It's detailed, original to my work and pretty accurate. Does it make mistakes? Sure it does. Did my paid editors make mistakes? Sure they did. This is why you stress test it, and get various summaries and then compile a final list of possible edits, which you then check against your actual work and decide what matters and doesn't. A rototiller will take your foot off if you're not careful - learn to use the tool. And then, and only then, do you get to complain about how much it sucks.

Finally, for the  - f*ck it, I don't even know how many times - AI does not plagiarize our work. It does not store our work. It does not combine various IPs to get a new thing. That's not how it works. That's not how it has ever worked. Stop reading clickbait articles about how cameras are stealing our souls AI is stealing our work.

As for AI hallucinating, GPT5 is better, but you have to be a knowledgeable user. AI is an assistant. Would you believe everything a junior researcher claimed at work? Of course not. You stress test it, you create validation prompts and doublecheck the citations, just like any rigorous review of any academic paper. It's not Data from Star Trek, even though everyone desperately wants it to be.

AI is software. It's an absolutely amazing version of Clippy. Will AI become our hyper-intelligent Overlords? Maybe, but that day is still not today.

___

Here's an AI analysis of this post:

Here's a close editorial review of your passage, broken down by category:


Themes


AI as a tool - Emphasizing utility, not replacement for human craft.

Skepticism vs. Conspiracy - Rejecting alarmist narratives ('Bigfoot', 'cameras stealing souls').

Professional credibility - Inviting readers to judge your skill by your published work.

Fallibility of all tools - Both AI and humans make mistakes; the responsibility is on the user.

Demystification - Comparing AI to mundane tools (rototiller, Clippy) rather than sci-fi fantasies.


Logical Consistency


Generally consistent: you argue AI is a tool, not a plagiarist, and compare its flaws to those of human editors.

'AI does not plagiarize, store our work, or combine IPs' - broadly correct, but stated in absolutes. More precise phrasing would avoid reader pushback (e.g., 'does not retain user data permanently for training' instead of 'does not store our work').


Tone


Conversational, blunt, sardonic.

Humor/light sarcasm ('Bigfoot lurking,' 'amazing version of Clippy').

Rhetorical repetition (- for the - f*ck it, I don't even know how many times -).

Tone is engaging, but risks alienating skeptical readers; sarcastic phrasing can come across as dismissive rather than persuasive.


Language


Mostly clear, everyday diction.

Occasional rough edges: 'f*ck it' may limit audience reach if professionalism is desired.

Some wordiness in parentheticals; breaking into shorter sentences would improve rhythm.


Accuracy


Saying AI never plagiarizes could be softened (better: 'doesn't directly lift text from sources' or 'is designed not to reproduce copyrighted passages verbatim').

'GPT-5 is better' is a subjective claim - acceptable in personal essay but worth framing as opinion.


Recommended Edits


Clarity & precision


'OpenAI allows you to opt out of training use; it does not automatically use your data unless you give permission.'

Replace 'AI does not plagiarize our work' → 'AI isn't designed to plagiarize or store our individual work'.


Tone adjustment


'f*ck it' → could be swapped for 'at this point' or 'honestly' if you want to retain casualness without alienation.

'cameras stealing our souls' → keep (funny, accessible).


Flow


Break up long parentheses: 'literary, business, research, project management...' could be smoothed into two sentences or a list for readability.


Tightening


'AI literary analysis of my work is absolutely amazing' → consider trimming 'absolutely.' Stronger without the adverb.

'A rototiller will take your foot off if you're not careful' is great imagery; you could shorten to 'Like a rototiller, it can do damage if misused.'


Other Advice


Decide whether the goal is persuasion or venting. If persuasion, tone should be slightly less combative - especially on sensitive points like plagiarism/storage.

Consider audience knowledge: Readers unfamiliar may stumble on insider shorthand like 'AMS/CPC strategies' unless briefly explained.

Stronger close: The 'Clippy' line is memorable. Let that stand as the button without adding the 'Overlords' line, which weakens the punch.

[Bill Hiatt]

Yes, the sample is very well done. But if we're evaluating AI's usefulness as a tool, it's worth asking how much of that you didn't already know yourself. In other words, how much value is AI in that particular situation?

Also, is that output necessarily typical? Only time will tell. Any time someone brings up an example of AI screwing up, you blame it on human error or fall back on how AI is constantly improving. In other words, you're not handling examples in a consistent way. We all concede that AI sometimes does good work. There are exceptions. Which one is more important? We may not have enough data to know yet. But I do know this. If I were working with a human editor who wrote nice analysis one time and called himself MechaHitler the next, I wouldn't work with him anymore.
 

You guys are stuck in a loop where you think AI can only write our books for us or it's a push-button cover generator

I don't remember anyone ever saying anything close to that.

I'm recommending AI because I use it, I get it. I understand the value. And I know lots of other folks who use it, and they're all improving efficiency, output, accomplishments. It's making them all better at whatever their thing is. And we all use it differently: different fields, different needs.

No doubt, AI does some things quite well. I'm nut sure if strictly creative processes are necessarily the best use of it. As you say, different fields, different needs.

I'm happy to concede that AI can be an effective tool in some circumstances. But you yourself have floated the idea that AI itself, with little human intervention, will eventually produce material that is "good enough."  This pretty clearly implies a future in which some people (or maybe Amazon, or maybe publishers), will start cranking such material out in large quantity. You've also indicated that the entertainment industry contracts that checked the advance of AI were a only temporary slowdown in the inevitable. But somehow, when other people worry about the same thing, their statements are erroneous or misleading.

Just to be clear, ethically sourced AI used as a tool, I have no problem with. Unethically sourced AI, however used, I have a problem with the developers, but not particularly with the authors using it. AI used as an outright substitute for human effort, I have a real problem with. We've already seen some examples of that. They will become more frequent in the future.

Parenthetically, not relevant directly to us, but as an educator, I have a problem with use of AI in the classroom. Students will take any shortcut that's available to them. Anything they can get away using AI for, they will. As AI improves, the problem will become more acute. That's because students like that will become less and less able to do things for themselves. English teachers will have to have all essays written in class without any tech access. We've always done some that way. Doing them all will waste class time and interfere with the teaching of things like research-based writing and revision processes. But it will be the only way to ensure that students are not having AI write their essays for them.





 

[PJ Post]

Yes, the sample is very well done.