AI edits rejected

[ronde]

Is using a calculator a crutch or is it a tool? I don't understand enough of trigonometry, but i can still use it to calculate the area expected to cover say the backyard for a pool. And it's been said the cosine tables in ships were often wrong and many which may have led to shipwrecks until they were corrected.

I'd LOVE to write my own 3D engine even if it were very simplistic but trigonometry i can't wrap my head around; If an AI or someone else wrote a basic one for me in a way that was usable, I'd probably use it, then write what i COULD myself for the rest of the framework, things I'm good at vs things I'm not. And if something was written that was bad I'd go over it with a fine tooth comb and try to fix it as best i could (or give up and delete the whole thing). Demanding it all be written fully by me when i can't understand the main most important parts though will just result in nothing ever being made.


But how good someone is for writing doesn't seem to be the issue. There's horrid stories by humans and horrid stories by AI, there's good stories by humans, and apparently good stories by AI. You're going to get good results or bad results, or bad results made into somewhat better results from external means (be it a friend giving feedback, paid editor or an AI, or skip it all and just hire a ghostwriter then get the credit while legally being in the right as owner of the work). But you can only bug your friend, mom, or someone else so many times before they just flatly refuse. And even if the AI were to do a lot of the heavy lifting, it's still trial and error so there's some growth there as you are unlikely to write a prompt, then print the result and never look at it. Maybe the personal growth is not as much as say using an oldschool 30's typewriter where you can only type 20 words a minute or it would jam up. Then again my mom used to say, if you wanted to be a good writer, read a lot of books, you'd innately pick up grammar structure plots and many other things that you can't learn elsewhere.

You just made my point. You apparently do understand enough trig to know you can use it to calculate some things beyond the scope of general math. That makes the calculator a tool you use to avoid using a set of trig tables. I also never said that tools were infallible, because no tool ever is. The results of using a tool depend entirely upon the understanding of the proper use of the tool by a knowledgeable user.

The ability of an author to write is indeed the question at hand. If I put your premise into a couple other endeavors, perhaps you'll understand.

I can play chess. I'm not a good chess player, but I have won a few games against other amateurs. Should I be able to bring a computer chess program to a match and let it tell me what moves to make against a more skilled player so that I can "win". To me, doing so would be cheating and would indicate to others that my goal is not to enjoy the challenge of playing the game. My primary goal would be winning however I can. Using a computer would not be using a tool. It would be using a crutch to make up for my lack of ability.

To use your own example, if I were to hire a ghost writer to translate my memoirs into a book, who really wrote the book? All I would have done is relate some experiences to a skilled writer. That's no different than asking a cabinet maker to make you a cabinet that will fit into a designated space and be able to hold a list of items. You generated the "prompt", but it was the cabinet maker who translated your "prompt" into something you can be proud of owning. Did you build it? No. Should you be proud to own it. Yes. Should you tell your friends you built it? No, because you didn't.

I would never recommend that any writer stop writing just because they don't get great ratings, but I would recommend they make an honest assessment of their ability and desire to become better by improving their skills at writing. AI is not the way to improve your writing. AI is a way to delude yourself into thinking that if you're not good enough, AI will make up for your lack of ability and low desire to improve on your own.

The best course of action in life is to enjoy doing the things you can do well, keep trying to improve on things you want to do better, and leave the things you can't or don't want to do to someone who can.

 

[SmilingLez]

I can play chess. I'm not a good chess player, but I have won a few games against other amateurs. Should I be able to bring a computer chess program to a match and let it tell me what moves to make against a more skilled player so that I can "win". To me, doing so would be cheating and would indicate to others that my goal is not to enjoy the challenge of playing the game. My primary goal would be winning however I can. Using a computer would not be using a tool. It would be using a crutch to make up for my lack of ability.


The best course of action in life is to enjoy doing the things you can do well, keep trying to improve on things you want to do better, and leave the things you can't or don't want to do to someone who can.

Using your chess example. If you play against the computer enough times and learn defense as well as offense when it comes to playing chess, you will become a better opponent and win more games (the chess computer becomes a tool, a tutor so to speak). The same would be true for AI. Learn about creating well-formed prompts (almost micro fiction) and generate enough AI stories (I am not talking about posting them as your own, so all AI haters get off of your high horse), examine and learn from the structure and dialogue; and in time your writing (structure) would improve. So in this case, AI storytelling is a tool to make human storytelling better. To reemphasize a previous point, AI stories are flat - they lack life.

YMMV

 

[yano2mch]

You just made my point. You apparently do understand enough trig to know you can use it to calculate some things beyond the scope of general math.

Even if i couldn't generate the numbers to use myself with any higher understanding of how they got those numbers; And only going off of the most minimalist of understanding. Beyond bumbling through to figure out a few circle or half circle estimations (which is some approximation of using pi or pi/2) i don't know. I do know i could get a circle drawn by telling BASIC to use sine/cosine, and do it faster by writing 8 points from 1 valued return rather than going through the whole gambit. I've no understanding of how tangent works or anything else.

I know my faults and limits, and complex math highschool/college math makes no sense to me. Don't make me use it, you will be disappointed by the results.

 

[ronde]

Using your chess example. If you play against the computer enough times and learn defense as well as offense when it comes to playing chess, you will become a better opponent and win more games (the chess computer becomes a tool, a tutor so to speak). The same would be true for AI. Learn about creating well-formed prompts (almost micro fiction) and generate enough AI stories (I am not talking about posting them as your own, so all AI haters get off of your high horse), examine and learn from the structure and dialogue; and in time your writing (structure) would improve. So in this case, AI storytelling is a tool to make human storytelling better. To reemphasize a previous point, AI stories are flat - they lack life.

YMMV

My question was not about playing against a computer. My question was should you use a computer to dictate your moves against another player. While there may be some benefit to watching a computer play against you, what you learn will be which move to make to counter the move you just made. You will never learn the logic behind what the computer does. It would be akin to following directions that only told you which way to turn without also including street names. You might get there, but if anything changed since you got the directions, you'll soon be lost.

I would argue that it may be possible to use AI to learn sentence structure, but what you'll learn is a very limited knowledge base derived from whatever resources the AI used to develop its responses. The beauty of writing is not copying what some software has "learned". The beauty of writing is taking an idea and using proper sentence structure to put that idea into your own words.

I've worked with robots for most of my career and it's amazing what they can do. Given enough time, it's entirely possible to program a robot to paint a copy of the Mona Lisa. From a distance, it would look pretty accurate. Close up, the infinitely variable techniques used by Leonardo da Vinci to give the painting "life" would be missing, just like the "life" of a story would be missing if you were to learn everything from AI. There comes a point where machines can not match human creativity.

 

[SmilingLez]

My question was not about playing against a computer. My question was should you use a computer to dictate your moves against another player. While there may be some benefit to watching a computer play against you, what you learn will be which move to make to counter the move you just made. You will never learn the logic behind what the computer does. It would be akin to following directions that only told you which way to turn without also including street names. You might get there, but if anything changed since you got the directions, you'll soon be lost.

I would argue that it may be possible to use AI to learn sentence structure, but what you'll learn is a very limited knowledge base derived from whatever resources the AI used to develop its responses. The beauty of writing is not copying what some software has "learned". The beauty of writing is taking an idea and using proper sentence structure to put that idea into your own words.

I've worked with robots for most of my career and it's amazing what they can do. Given enough time, it's entirely possible to program a robot to paint a copy of the Mona Lisa. From a distance, it would look pretty accurate. Close up, the infinitely variable techniques used by Leonardo da Vinci to give the painting "life" would be missing, just like the "life" of a story would be missing if you were to learn everything from AI. There comes a point where machines can not match human creativity.

I thought you were talking about using a tool versus a crutch. AI right now for most people is a crutch when it comes to storytelling. That is a shame.

 

[yano2mch]

I thought you were talking about using a tool versus a crutch. AI right now for most people is a crutch when it comes to storytelling. That is a shame.

I'm reminded; You usually need at least two people to run a business. You have the creative head, who has ideas and is a people person, and can make the product you're interested in. Then you need the businessman who can sell it, who can advertise it and get it marketed.

Very very few people can do both jobs. Those that can, are likely very lucky.

I've said before, I'm an RPer, i can write and do it very well. But i can't do long segments. So if i end up using AI so any kind of story or book can push past 3 paragraphs, then so be it (plus there's no argument over who owns the final product). And in the process I'll still have put my heart and soul into it; Even if an AI happened to help.

 

[ElectricBlue]

Now. Are you going to argue that a model trained with say with image of Darth Vader and Harry Potter text, would be stealing from Lucas Films and JK Rowling? Or are you saying an output using any of that content even if it isn't anywhere near close to any instances that they'd made is considered stealing? Or converting text from one language to another is also stealing?

You've missed the point that the AIs have been trained on the copyright content of authors and artists (as well as older content in the public domain, but that's a form of cultural theft), but with no royalties or payment made to those artists. That's theft, pure and simple.

That's the issue with the LLMs, not whether or not you can copyright the outputs. It's the input that matters, for writers, musicians and artists, the original creative thinkers. That's why most people here don't give a toss about those who say, "But AI is only a tool, it's okay to use it." No, it's not, it's theft.

I can't play an instrument, but it would never occur to me in a million years to go use some musical AI thing, stick in a few prompts, and call myself a musician. Writing is no different. If you can't do it yourself (you know, by learning grammar, how to spell, how to know your own style, etc), you're kidding yourself if you think you're a writer.

 

[SmilingLez]

I can't play an instrument, but it would never occur to me in a million years to go use some musical AI thing, stick in a few prompts, and call myself a musician. Writing is no different. If you can't do it yourself (you know, by learning grammar, how to spell, how to know your own style, etc), you're kidding yourself if you think you're a writer.

So does that mean if you wanted to create music and had a Mac Pro you would not use Apple logic Pro and or FL Studio?

 

[yano2mch]

You've missed the point that the AIs have been trained on the copyright content of authors and artists (as well as older content in the public domain, but that's a form of cultural theft), but with no royalties or payment made to those artists. That's theft, pure and simple.

Right... and any story involving a robot should have had money going to Isaac Asimov, anything involving space and rockets we should have had to pay homage to Jules Verne, and anytime fireworks is involved the Chinese dynasty back in the 1300's should get money too i suppose.

If i draw a car perhaps i should have to pay royalties to Buick, or if i play on an instrument Yamaha needs payments too; if i talk about Little Mermaid probably give money to Brothers Grimm, though Disney would prefer to sue and say it's their thing even if they took it from public domain.

Again, AI's are trained and weights to find PATTERNS, not copyrighted works. The copyrighted work may nudge the model in a small way but when you get the model as weights there isn't a single paragraph that was copyrighted, only the likelihood word X is followed by word Y, with a back context of Z.

Though i do suppose you are the type of person when you take photos, you are sure to cover EVERYTHING with a sheet so nothing can be identified from any one brand because you might have to pay them for having it inside your image and breaking copyright/trademark.


Copyright is very broken, currently it pushes for corporations to own everything and you to own nothing and you have to rent pay-per-view for any content you want. Or at least that's the future they'd love. And if you start to make anything interesting or challenging, they will sue you to the ground or buy you out, thereby keeping their monopoly.

edit: Though i do suppose i could go into the argument you are a thief yourself, because your brain was trained on copyrighted material and anything you come up with is some subset result of your life experiences, which is from copyrighted material...

 

[ElectricBlue]

So does that mean if you wanted to create music and had a Mac Pro you would not use Apple logic Pro and or FL Studio?

I'm not a musician, so I have no idea if those things are AI tools or not. If they've scraped original content for their algorithms (whatever they might do) then the same applies as for LLM for written content - copyright theft is copyright theft, so far as I'm concerned.

 

[SmilingLez]

I'm not a musician, so I have no idea if those things are AI tools or not. If they've scraped original content for their algorithms (whatever they might do) then the same applies as for LLM for written content - copyright theft is copyright theft, so far as I'm concerned.

It's sort of an in-between. Not really AI but more algorithmed then for example, Word (before any AI junk)

 

[yano2mch]

It's sort of an in-between. Not really AI but more algorithmed then for example, Word (before any AI junk)

Could be comparable to Casio keyboards, which includes like 30 sets of drum beats and patterns, and sub-patterns when you hit a chord. Along with sound fonts shoving dozens of instruments in 2Mb of space.

But can't have obviously commonly used patterns... all that evil musician stuff... Probably better to bang pots together.

 

[ronde]

I thought you were talking about using a tool versus a crutch. AI right now for most people is a crutch when it comes to storytelling. That is a shame.

It does seem that every one of these threads contain some statement about using AI to make a story better and there should be nothing wrong with that. "It's still my idea and my words" is another common statement.

That makes AI a crutch for those writers, just as you say.

My point about playing chess and a robot painting a picture is both use some form of computer "intelligence" in order to function. While not AI in the way we currently understand it, a computer program that can project ahead several possible chess moves and then decide which one is the safest is not all that different than AI scanning the internet for acceptable sentence structures and picking the one it deems to fit in a particular sentence. The chess computer is just confined to a much narrower dataset.

Likewise, the robot program appears to be relatively simple on the surface. What the robot can not do is duplicate the actual brush strokes of a human painter because there is no predictability to them, just as there is no predictability to what a good writer puts down on paper, using a typewriter, or a word processor. The method of recording the idea is a tool. The inspiration that governs those brush strokes or word patterns is not something AI can do, at least not currently.

Letting computing power make the decisions about sentence structure or brushstrokes or anything else that requires human creativity makes AI a crutch and not a tool.

 

[AngelaSaxon]

Ok can people tell me, are there actually lots of people whose writings are being falsely flagged as AI in some way here?

Or just people who ARE using AI but feel like THEIR use should be acceptable?

There may be arguments and discussions to be had in both cases, but they seem like very different situations to me.

So if people have clearly addressed that question and the answer should be obvious to me then I apologise, but I only read a dozen or so messages in this thread, not the entire thing, and the parts I read all seemed to be people who HAD used AI being flagged, but some responses writing giving advice as if people were being flagged for 'writing too well' or something...?

 

[ElectricBlue]

Ok can people tell me, are there actually lots of people whose writings are being falsely flagged as AI in some way here?

Or just people who ARE using AI but feel like THEIR use should be acceptable?

I suspect both. There's are dozens of threads in the Authors Hangout about the AI scourge and how the site appears to be managing it.

The site hasn't disclosed what AI detection processes they've adopted, but suffice to say there's an unknown percentage of writers who appear to be getting false negatives. However, judging by the reduced number of threads now, compared to this time last year, either the site has got a level of sophistication in their "whatever it is they're using", and this number is decreasing (but hits still occur), or there's less AI assisted content coming through. No-one here sees the actual numbers, so this comment is based on anecdotal evidence - fewer threads.

There may be arguments and discussions to be had in both cases, but they seem like very different situations to me.

They are, but the site's policy is clear - don't submit stories written with AI assistance, they'll be rejected.

Most writers in the AH support the policy, judging by responses to threads. There are a few who argue for it, basically saying, well it's here, you should just get used to it - but they won't get published if they openly declare what they're doing.

So if people have clearly addressed that question and the answer should be obvious to me then I apologise, but I only read a dozen or so messages in this thread, not the entire thing, and the parts I read all seemed to be people who HAD used AI being flagged, but some responses writing giving advice as if people were being flagged for 'writing too well' or something...?

Check out more threads in the AH.

 

[AngelaSaxon]

I suspect both. There's are dozens of threads in the Authors Hangout about the AI scourge and how the site appears to be managing it.

The site hasn't disclosed what AI detection processes they've adopted, but suffice to say there's an unknown percentage of writers who appear to be getting false negatives. However, judging by the reduced number of threads now, compared to this time last year, either the site has got a level of sophistication in their "whatever it is they're using", and this number is decreasing (but hits still occur), or there's less AI assisted content coming through. No-one here sees the actual numbers, so this comment is based on anecdotal evidence - fewer threads.

They are, but the site's policy is clear - don't submit stories written with AI assistance, they'll be rejected.

Most writers in the AH support the policy, judging by responses to threads. There are a few who argue for it, basically saying, well it's here, you should just get used to it - but they won't get published if they openly declare what they're doing.

Check out more threads in the AH.

Thanks.

It's a fairly abstract question for me, I have no interest in using AI in any writing.

I guess I wonder if there really ARE many people being flagged as AI who haven't used AI at all. But hey maybe, I have no idea how good AI detection is.

Personally I would err on the side of flagging even minor use rather than letting stuff slip through, because I prefer to read bad writing from actual people over whatever passes for 'good' when congealed by a more elaborate version of Eliza...

 

[yano2mch]

Thanks.

It's a fairly abstract question for me, I have no interest in using AI in any writing.

I guess I wonder if there really ARE many people being flagged as AI who haven't used AI at all. But hey maybe, I have no idea how good AI detection is.

Personally I would err on the side of flagging even minor use rather than letting stuff slip through, because I prefer to read bad writing from actual people over whatever passes for 'good' when congealed by a more elaborate version of Eliza...

I'm against whole works being made by AI. Prompt 'write me a best selling novel' etc. that is obviously disallowed.

But a lot of tools disclosed as AI or not, are more writing assistants. Like a live thesaurus or the like. I'd argue the rules are way too strict. You shouldn't get flagged as AI writing, for moving a comma or having spell checking from an AI. But these too are being flagged.

I'd say AI detection should likely be used a lot more in... oh say homework, college thesis's and the like to make sure kids aren't cheating, and then of course other professions (like a lawyer drafting an argument, filled with fake case numbers and fake arguments because it hallucinated it wholesale).

Though I've heard in school AI detection being overkill too. A friend's kid's entire homework assignment was rejected as 'being AI', then when prompted where it was 'at the beginning'; Which literally the definitions for what the subject was about, likely copied word for word (to be correct) from a dictionary or assignment paper, before moving onto the actual homework of the topic.

It's going into a inverted justice system, where you are willing to destroy 100 good/great stories/works in exchange for 1 really bad work of dubious quality because it has more errors and thus can't be AI.

 

[AngelaSaxon]

I'm against whole works being made by AI. Prompt 'write me a best selling novel' etc. that is obviously disallowed.

But a lot of tools disclosed as AI or not, are more writing assistants. Like a live thesaurus or the like. I'd argue the rules are way too strict. You shouldn't get flagged as AI writing, for moving a comma or having spell checking from an AI. But these too are being flagged.

I'd say AI detection should likely be used a lot more in... oh say homework, college thesis's and the like to make sure kids aren't cheating, and then of course other professions (like a lawyer drafting an argument, filled with fake case numbers and fake arguments because it hallucinated it wholesale).

Though I've heard in school AI detection being overkill too. A friend's kid's entire homework assignment was rejected as 'being AI', then when prompted where it was 'at the beginning'; Which literally the definitions for what the subject was about, likely copied word for word (to be correct) from a dictionary or assignment paper, before moving onto the actual homework of the topic.

It's going into a inverted justice system, where you are willing to destroy 100 good/great stories/works in exchange for 1 really bad work of dubious quality because it has more errors and thus can't be AI.

I don't know, people have mostly managed to move commas and check spelling without the use of AI for decades at least and I'm sure people can manage to do so again if that's REALLY what's going on.

 

[yano2mch]

I don't know, people have mostly managed to move commas and check spelling without the use of AI for decades at least and I'm sure people can manage to do so again if that's REALLY what's going on.

I agree. But some things people are terrible about and need a second set of eyes. Also for decades you've have (paid and unpaid) editors, teachers, parents and others to look for and mark and point out errors vs it all being you, or even ghost writers who are probably specialized for that type of task. There's several times I've stared at something and couldn't see it in front of me, and someone just points it out and 'oh now i see it!'. Like double words in in your sentence (often at the wraparound on the page/paragraph).

It sounds more like they want you to put in your first draft of something with no outside input or help at all. This is a more 'this isn't real art because you didn't draw it on pen and paper' argument they had when people started using digital tablets. Now everyone who does art mostly does those and physical art is a pasttime. And remember with typewriters before you had electronic ones, you had to use white-out when you made mistakes. A text work can't be real if you didn't physically type and fix every little mistake yourself afterall right? No CTRL-Z or hitting backspace, or writing an entire line and making fixes before it actually outputs.