AI vs. Professional Authors Results

http://mark---lawrence.blogspot.com/2025/08/the-ai-vs-authors-results-part-2.html

by biffles • 11 months ago

94 95 11 months ago

mark---lawrence.blogspot.com

Aeolun4• 11 months ago

Personally I’m having a blast reading AI generated fiction. As long as the direction is human, and often enough corrected to keep the minor inconsistencies out, the results are pretty good.

For me it’s no different from generating code with Claude, except it’s generating prose. Without human direction your result ends up as garbage, but there’s no need to go and actually write all the prose yourself.

And I guess that just like with code, sometimes you have to hand craft something to make it truly good. But that’s probably not true for 80% of the story/code.

antihipocrat• 11 months ago

Who's voice are you using when adding your hand crafted prose? Mimicking the style of the 80% or switching to your own?

Perhaps I'm a Luddite, or just in the dissonance phase toward enlightenment, but at the moment I don't want to invest in AI fiction. A big part of the experience for me is understanding the author's mind, not just the story being told

add-sub-mul-div• 11 months ago

A big part of the experience for me is understanding the author's mind, not just the story being told

AI content is really exposing how people fall into a group that does go further than the surface text into deeper layers of context/subtext, and a group that doesn't.

Aeolun4• 11 months ago

I think there’s only a very small subset of fiction that uses the prose to that extent. Much like code really. If you are writing original algorithms you cannot use the LLM. If you are just remixing existing ones, it becomes a lot more useful.

Also, I guess I missed the brunt of your question, though the answer is similar. Most voice works for most characters. There’s only so many ways to say something, but occassionally you have to adjust the sentence or re-prompt the whole thing (the LLM has a tendency to see the best in characters).

CuriouslyC• 11 months ago

Plot twist, people who do first drafts and structural edits with AI can still do line edits and copy edits by hand for personal voice (and you have to anyhow if you want the prose to be exceptional).

sram1337• 11 months ago

I would love to read some of this. Where do you find AI generated fiction?

Aeolun4• 11 months ago

https://www.royalroad.com/fictions/search?globalFilters=fals...

Though I’ll admit I can’t speak to the quality of that except my own stuff (which I’m naturally predisposed to like).

This was my attempt at fully AI generated (though edited by human):

https://www.royalroad.com/fiction/101072/inherited-wounds

deadbabe• 11 months ago

The only AI generated fiction I read is stuff I create on the fly. Why would you read AI generated fiction made by other people when it’s the same as reading regular fiction?

jay_kyburz• 11 months ago

Just wanted to be first to say I'm a huge fan of Mark Lawrence. I didn't know he blogged! Now I'll go actually read the post.

dwd3• 11 months ago

Big fan of Janny Wurts, particularly the Empire series she co-wrote with Raymond E Feist. Very surprised she was outed as an AI, though she is the only author I've ever had to pull out the dictionary as she used a word I had never seen and wanted to be sure of the meaning in the context.

npinsker• 11 months ago

I was very (wrongly) confident she was AI. I thought her prose was bland and repetitive and lacking melody, and she was the only human author to explicitly mention demons in the first paragraph, which is something that happens when you prompt an AI for a demon story.

The twist had potential, but to me wasn’t executed as well as it could have been. I was hoping for some stronger irony — e.g. if the mainlanders had pushed for the bridge, but not the islanders; or a sentence about how the demons were very surprised, but nonetheless went on to take over the world before they knew what had happened. As it is, it felt underdeveloped and slightly uncanny in an AI-like way.

kristopolous• 11 months ago

ai fiction really shines in baffling surreal prompts that it tries hard to satisfy. Here's an example:

"let's write a story where donald trump is giving a speech to a crowd as people slowly discover he is secretly a northern red oak tree in a human suit to the shock of fans and reporters! The tips of his fingers become branches as he tries to deny it as wildly impossible meanwhile his human disguise continues to fail"

After some back and forth, here is what I got:

https://9ol.es/md/trump

No human would write something that crazy...

Ancapistani• 11 months ago

So... Kafka?

akoboldfrying• 11 months ago

I applaud OP's transparency and willingness to call this result what it is.

xenotux2• 11 months ago

I think this is similar to AI generated images: it puts a new creative tool in the hands of people who might have had good ideas, but didn't have a mastery of the medium. In that respect, it's cool: if you had a great idea for a sci-fi story but no talent for writing, and if an LLM let you realize your vision, that's neat. It has some negative externalities for the craftsmen, but overall, more creativity is hardly a bad thing.

The real problem is that the most lucrative uses of the tech aren't that. It's generating 10,000 fake books on Amazon on subjects you don't care about. It's cranking out SEO spam, generating monetizable clickbait, etc.

gtowey• 11 months ago

And they haven't even gotten around to adding advertising into them yet! Imagine when chat assistants subtly steer you towards certain products. Would you even know it was manipulating you?

perching_aix2• 11 months ago

They kind of did though. Thing tries to make me generate diagrams every step of the way (as a kind of feature demo), even though they're rarely ever a good idea, and even when they are, the pictures and diagrams generated are useless.

thwarted4• 11 months ago

I think this is similar to AI generated images: it puts a new creative tool in the hands of people who might have had good ideas, but didn't have a mastery of the medium.

Reading this sentence reminded me of the classic HN position of "ideas are worthless, what matters is the execution", usually mentioned in the context of an "ideas person" looking for their "technical cofounder" and the ideas person thinking they deserve at least 50%, often more, of the ownership of what would be built because without them there'd be no idea.

if you had a great idea for a sci-fi story but no talent for writing, and if an LLM let you realize your vision, that's neat.

If your "vision" is only the "idea for a sci-fi story", is that really a vision? Good books leave the reader changed/influenced in some fashion, through the way the idea is presented and developed over the course of the story, not just from a blurb on the book jacket.

overall, more creativity is hardly a bad thing.

Is coming up with an idea for a for a sci-fi story the meat of creative act such that that flooding the market with ideas counts as an increase in creativity overall?

vunderba3• 11 months ago

Agreed. Have a read through the 10 dragon stories - the majority of them are rich in spices, but bereft otherwise.

LLMs seem to revel in throwing layer after layer of decorative paint in the hope that people will fail to notice that they're not actually painting anything.

As a writer, the best advice that I can give is to build your house upon the rock and not upon the sand.

hyperadvanced2• 11 months ago

I also think that just having “an idea” isn’t exactly the same as increasing net creativity. Oftentimes with art, the impact of something truly creative results in changing the parameters of the medium or genre itself, not merely sticking to the script and producing a new work in the style of X. If you take, for example, J Dilla and his impact on hip-hop, the fact that there’s an entire subgenre or two focused on some of his hallmark innovations (micro-rhythm/wonky beats, neo-soul, lofi sampling/creative use of samples) speaks to that kind of “real” creativity. I frankly think that kind of genre-bending is possible with the use of LLMs, but if you just say, “here’s my story idea, make it so”, without any eye towards the actual technique or craft, you won’t be getting the next Blood Meridian out of it.

d0100• 11 months ago

If your "vision" is only the "idea for a sci-fi story", is that really a vision?

We have art, games and movie directors

AI just enables anyone to be a "director", but most people can't direct anything worthwhile

colechristensen• 11 months ago

Having tried to use AI as a technical cofounder (in small scale experiments), nah it's not there yet. It's more like having a herd of interns. With close supervision there can be a lot of productivity but without a good deal of taste and expertise it's just going to be garbage unless you're very lucky.

xenotux2• 11 months ago

I think that's an odd way of viewing creativity: unless you can pull off the whole thing, you're not really creative/

What about people who are not native speakers? Who are dyslexic? Do we deny them the spark of creativity because they can't write perfect prose without help? Heck, what about most sci-fi writers? Their editors often do a lot of heavy lifting to make the final product good.

If you have a killer idea for a meme or a really clever concept of a four-panel comic strip, but don't know how to use Photoshop or can't draw very well, is it a sin to ask a machine to help? Is your idea somehow worthless just because you previously couldn't do that?

I'm not disputing that a lot of people don't use these tools this way. In fact, that was exactly my point. If your "idea" is to crank out deceptive drivel, I'm not defending that.

andrewflnr5• 11 months ago

more creativity is hardly a bad thing.

Why, exactly, is creativity good? What is the benefit, and to whom? Does that benefit survive the interposition of genAI? I'm doubtful, either for the reader or the craftsmen.

entropyneur• 11 months ago

Wondering if anyone had success with this yet. I have several ideas for poetry and prose that I don't have the skill to pull off. I periodically plug them into new models and so far all the results have been completely unsatisfactory.

bsder• 11 months ago

Perhaps what this is pointing out is that a lot of writers of the genre of "fantasy" produce mostly formulaic, trope-laden piles of crap that AI is pretty good at mimicking?

This is neither new nor news. "The Well-Tempered Plot Device" is almost 4 decades old (see: https://news.ansible.uk/plotdev.html).

It does suggest that publishers might want to screen new writing with a quick "Did AI write this?" and only publish the ones where it is obvious to humans that AI did not write it.

spondylosaurus• 11 months ago

I thought it was interesting/telling (but maybe not surprising) that the AI-generated stories scored the highest according to reader rankings, yet pinged to me as immediately flat and generic. But I really liked the idiosyncrasies of a few of the human-authored entries!

exmadscientist3• 11 months ago

Yes, the human-written stories that I guessed wrong about were the ones that seemed to have nothing to say. When the plot's stereotypical and trite, there's no subtext (difficult to do in flash, but not impossible; some can do it well), and it scans like anyone could have written it anywhere or anytime, well, that looks like AI.

(In that vein I am baffled how anyone could think the fourth story, especially, was anything but AI.)

(And, as well, the seventh story is interesting because it reads, to me, exactly like someone who's used to writing something longer trying to write flash. It doesn't land anything, it doesn't conclude, but it looks like if it had about twice the length it might be interesting. And it's got some dissonance from breaking with the usual demon-bargaining template. So I pegged that as human. Oops!)

andrewflnr5• 11 months ago

"The Well-Tempered Plot Device" is almost 4 decades old

Yeah, we've moved forward a ways in the last 4 decades, or the top of the market has, at least. That was a fun read, though.

spongebobism2• 11 months ago

Quoted in the well-tempered plot device:

Our relationship / is beautiful / because / it is ours / because / it relates / to us.

Indistinguishable from Rupi Kaur. There is nothing new under the sun.

vunderba3• 11 months ago

The story plot was "Meeting a dragon". As both a human and a writer, challenge accepted:

Long ago, there lived a golden dragon whose fractal-like scales gleamed in the glow of the morning in her cave. She was known for her kindness, and many came not with sword or spear, but with humble requests - for you see, it was widely believed that the mystical scales of a dragon would heal illness, cure ailments, and provide fortune.

One such visitor timidly looked up at her great shining body and beseeched, "Oh glorious dragon, might I have a single scale?"

Of course, the dragon replied warmly. She delicately, almost lovingly, with a slight twinge, used a single claw to prise off a single golden scale, leaving a dull patch.

Over the eons, more and more people would come as supplicants. The scales were used for good luck, for warmth, to ward off evil, as the draconic equivalent of a rabbit's foot.

In the end, the poor dragon was stripped bare - the fire from her burning furnace now showed clearly through to her patchwork, sensitive, and naked skin.

When winter came, she huddled in the cold darkness. And still, when a peasant would come asking for a scale - just one, a single scale nothing more, she would not refuse. In her eternal generosity she would carefully break off another. This time it took longer to find one left upon her body, as the humans had stripped her bare like a tree come winter.

Then thus came a knight. "I'm sorry, good sir, but I have no scales left to give," she said pitiably.

"Why, your scale was a choking hazard and wasn’t labeled not for ages under 5! Prepare for a class-action lawsuit and also to be impaled upon a lance."

The End.

I'll pretend I intended it as a parable of the destructive nature of mass tourism or something something Lorax something something truffula trees.

eulgro• 11 months ago

Thanks, I love it.

taneq• 11 months ago

No good deed goes unpunished!

andrewflnr5• 11 months ago

For me, the part about "fractal-like scales" would have flagged the author as either an AI or just kind of a pretentious dummy, because that makes very little sense. "Then thus" would actually lean me toward human, because LLMs are usually better at grammar than that. :D

lelanthran2• 11 months ago

I think that humans do executive decision making better than LLMs.

So, yeah, your "Meeting a dragon" story was about a single point - an attempt at a humourous twist ending; you built your story around that.

My approach would be having the Dragon in the title be a metaphor, for something powerful, dangerous and scary:

1. Comedy: new g/friend meeting MiL (the Dragon) for the first time

2. Thriller: guy finds out what his wife (The Dragon) really is like, basically the plot to Gone Girl

3. Drama: An alcoholic anthopormorphising the addiction (The Dragon) as an uncontrollable beast within himself

4. Historical: An author examining the events of the Tuskagee Syphilis experiment in the legislated-racism (The Dragon) period of the time.

5. SciFi: "The Dragon is a Harsh Mistress" (enough said)

6. Action: The dragon is a legendary elder of a mystical martial arts sect ("Enter the Dragon" stuff)

7. Fantasy: An actual, literal Dragon!

This is before I've considered a single line of plot, a single character, or character motivation. It's before I considered tone and presentation (Narrative? First-person narration? Dialog-driven narration?)

I mean, before getting to actual plot, characters, setting, tone, etc ... I've already got my message figured out. That's executive decision making.

The LLM will not, when given the directive "Write the story 'Meeting the Dragon'" perform any executive decision making. You have to baby it through every step. Basically micromanage it.

Abdii430• 11 months ago

https://www.facebook.com/profile.php?id=61579582377159

thedevilslawyer4• 11 months ago

We're at a key milestone in the IP wars; IP was inducted into our civilization to protect the few who had the ability to create.

This idea is well past it's due date. We should move to a liberal IP regime, with copyright strictly reduced to 7-10 years, with all works then entering public domain. Our society will universally thrive with the abundance that will come.

I understand and empathize that a class of vocations today will go away, but so did lamplighters. The roles may become extinct; but we will endure as a people.

HPMOR• 11 months ago

Why is it always taken as a conclusion that humans will persist? Lots of species go extinct, and it is not clear that won’t be true here as well.

thedevilslawyer4• 11 months ago

We won't go extinct because of the removal of an arbitrary law around monopolizing ideas.

rwyinuse• 11 months ago

I'm not sure people left without a job will endure, at least if they're unfortunate enough to live somewhere without social security. The key question is whether AI will somehow create enough new jobs for those let go (I doubt that), or just cause a massive unemployment.

thedevilslawyer4• 11 months ago

Indeed, I'm proposing that we figure out a solution for this. That would be better than bearing the cost of knowledge not being open and universal.

ordinaryradical• 11 months ago

There are hundreds of years of art already in the public domain and society is not “universally thriving?” Why would adding a couple more recent years change that?

Seems very utopian magic thinking to me.

thedevilslawyer4• 11 months ago

The vast majority of the works have been created in the last 100 years. This is where the "intellect" of IP lies, and that's what's locked behind copyright. Opening these up would absolutely affect universally thriving.

You can't live modern life with knowledge as of 1900.

voidhorse3• 11 months ago

Interesting study. I think the use of AI boils down to this: is the product independent of the process and the context? Or is it dependent on it in some way? I think, when it comes to art, the latter is truer than the former, and most use of AI in creative fields is predicated on trying to convince people to engage with art in an extremely shallow way (art strictly as soulless, time filling entertainment).

When an author writes a novel, the novel does not exist in a vacuum. The author's persona and the cultural exchange that emerges around the text also becomes an important part of the phenomenal status of the the work and its cultural recognition. Even when an author remains pseudonymous and makes no appearance, this too is part of the work.

If an author uses AI as a tool and takes care to imbue the output with artistic and personal relevance, it probably will become an art object, but its status may or may not be modulated by the use of AI in the process to the extent that the use does/doesn't affect people's crafting interpretations of the work, or the author's own engagements. Contrarily, AI generated work that has close to no actual processual involvement on the part of the author will almost always have slop status, just because it's hard to imagine an interpretive culture around such work that doesn't at some point break down in the face of the inability to connect the worm with other cultural touchstones or the actual experience of a human being. Maybe it could happen, but if it did, at that point the status of the work is till something different in so far as it would be a marker not of human experience, as literature traditionally has been, but something quite new and different literature-cum-hypermarket(we already had mass market) product.

waltbosz• 11 months ago

There is an Asimov story called "Someday" in which a toy computer called a Bard generates random fairy tales and reads them to children.

In the story two children try to hack their Bard, to make it tell more interesting modern stories, by feeding it a new vocabulary of modern words. In the end, it just generates the same old fairy tale plots using the new words it has learned.

I really feel like that story embodies today's AI generated stories. I've tried to get ChatGPT to generate original fairy tales and whatever plot prompt I give it, it spits out what is essentially the same dull story every time.

I always enjoy spotting a good anachronism in a sci-fi story (societies with space travel but still use typewriters), but this is a case of really spot on prediction.

thenoblesunfish• 11 months ago

Is this why the Google product (now just called Gemini) was called that?

godelski5• 11 months ago

I don't know if anyone has officially said so but there was a public statement about it being chosen as a reference to story telling (Celtic language). So could be for similar reasons or could be a reference. Not surprising considering how famous the story is and how famous Asimov is. But maybe someone else knows more definitively.

godelski5• 11 months ago

  > it just generates the same old fairy tale plots using the new words it has learned.

I think you're leaving out the best part! I don't want to spoil it, it's a short story. Classic trope, but still. Story here[0]

On another note, as an avid SciFi lover I have always found it interesting that in books, movies, and shows there have been many machines that talk and do complex tasks yet no one ever thought they were alive. Just take Star Trek. The simulations in the Holodeck are highly realistic and intended to mimic real humans. Or even the computer is able to speak and write code as requested. Far more advanced than our systems today. There's even that famous episode in TNG with Data where they are questioning if he is actually alive or not. Not such an easy thing but yet every viewer probably thought he was and recognized the difference between him and the computer and Holodeck[1]. Though my favorite version of that question is in Asimov's The Positronic Man (basis of the movie Bicentennial Man and yes, Asimov is why Data has a Positronic brain). These are fiction, but I find this so interesting. I feel like our LLMs look much more like the computer from Star Trek than the Holograms let alone Data. Yet, I think there's a lot of disagreement about the level of intelligence of these systems and it makes me wonder why someone would say the computer in Star Trek isn't intelligent but the LLM is (I'm sure there's retconning too).

[0] https://nyc3.digitaloceanspaces.com/sffaudio-usa/mp3s/Someda...

[1] Well there is Voyager. And that episode from TNG. But go read [0] ;)

famouswaffles4• 11 months ago

Because it's fiction and the Author is God.

In Star Trek, the computer is framed as an appliance. It's the ship's operating system. The characters treat it like a highly advanced Alexa. They issue commands ("Tea, Earl Grey, hot"), ask for information, and expect a transactional response. No one ever asks the computer, "How are you feeling today?" because the narrative has established it doesn't have feelings. It's a tool, and we, the audience, accept this premise.

In contrast, the entire point of Data's character is to question the line between machine and person. The episode you mentioned is a courtroom drama specifically designed to force the characters (and the audience) to see him as a sentient being with rights. His "positronic brain" is the magical Asimovian hand-waving that signals to the audience: "This one is different. Pay attention."

'The Author' could have easily positioned the computer or the holodeck in a similar manner and people would agree it was sentient. Or Star Wars droids could easily be given more of this kind of weight than they are currently given.

It's one thing to read a fictional story about a fictional technology and assume the position and framing the God is pushing you to, it's another thing entirely to have the technology in your hands and play around with it.

Agraillo• 11 months ago

I've tried to get ChatGPT to generate original fairy tales and whatever plot prompt I give it, it spits out what is essentially the same dull story every time.

Not a universal solution, but a working method to get at least sometimes interesting results. You should use it as a co-authoring tool by following these hints: treat this as a dialogue ("let’s create interactively, you and me…", "create a first sentence of a fictional story", ...), where you act like a semaphore for the continuation—judging the current output and either correcting it or suggesting the next step (which can be brief and expanded later by the LLM). Finally, try to suggest unexpected constraints. This can be effective because when your constraint contains a set of words rarely seen together in training data, the output becomes somewhat random but still at least partially grounded in reality.

An example from one of my old conversations with Llama 3.1:

User: Create a sentence from a fictional book containing the words crazy, cowboy, and gadget.
Llama: In the wild west town of Crazy Horse, a notorious cowboy named Buckshot Bob unveiled his latest gadget — a mechanical horse that could gallop at breakneck speed.

dwd3• 11 months ago

The best fantasy/sci-fi literature involves a lot of world building.

For some, the world building came first and the stories were an offshoot of that.

Tolkien needed a world and stories to bring life to the languages he was inventing.

Raymond E Feist's Midkemia was a massive collaborative effort for a RPG world. He has stated: "I don't write fantasy; I write historical novels about an imaginary place. At least that's how I look at it."

This is what you won't see AI doing...yet.

antisthenes• 11 months ago

World building...is just having a really large context window.

ethan_smith• 11 months ago

Multi-agent LLM systems with persistent memory are actually making significant progress on world-building coherence, maintaining consistency across thousands of interactions while incrementally developing complex fictional universes.

unignorant4• 11 months ago

Here are my notes and guesses on the stories in case people here find it interesting. Like some others in the blog post comments I got 6/8 right:

1.) probably human, low on style but a solid twist (CORRECT) 2.) interesting imagery but some continuity issues, maybe AI (INCORRECT) 3.) more a scene than a story, highly confident is AI given style (CORRECT) 4.) style could go either way, maybe human given some successful characterization (INCORRECT) 5.) I like the style but it's probably AI, the metaphors are too dense and very minor continuity errors (CORRECT) 6.) some genuinely funny stuff and good world building, almost certainly human (CORRECT) 7.) probably AI prompted to go for humor, some minor continuity issues (CORRECT) 8.) nicely subverted expectations, probably human (CORRECT)

My personal ranking for scores (again blind to author) was:

6 (human); 8 (human); 4 (AI); 1 (human) and 5 (AI) -- tied; 2 (human); 3 and 7 (AI) -- tied

So for me the two best stories were human and the two worst were AI. That said, I read a lot of flash fiction, and none of these stories really approached good flash imo. I've also done some of my own experiments, and AI can do much better than what is posted above for flash if given more sophisticated prompting.

codechicago277• 11 months ago

I had similar results, and story 4 is so trope heavy I wonder if it’s just an amalgamation of similar stories. The human stories all felt original, where none of the AI ones did.

unignorant4• 11 months ago

I'm not sure I agree that the human stories felt original. I was pretty unimpressed with all of the stories except maybe 6, and even that one dealt in some common tropes. 5 had fewer tropes than 6 (and maybe as a result of that received the highest average scores from his readers) but I could tell from the style it was AI

biffles• 11 months ago

Could you expand on your point re more sophisticated prompting?

I have found it hard to replicate high quality human-written prose and was a bit surprised by the results of this test. To me, AI fiction (and most AI writing in general) has a certain “smell” that becomes obvious after enough exposure to it. And yet I scored worse than you did on the test, so what do I know…

unignorant4• 11 months ago

For flash you can get much better results by asking the system to first generate a detailed scaffold. Here's an example of some metadata you might try to generate before actually writing the story: genres the story should fit into; pov of the story; high level structure of the story; list of characters in the story along with significant details; themes and topics present in the story; detailed style notes

From there you have a second prompt to generate a story that follows those details. You can also generate many candidates and have another model instance rate the stories based on both general literary criteria and how well the fit the prompt, then you only read the best.

This has produced some work I've been reasonably impressed by, though it's not at the level of the best human flash writers.

Also, one easy way to get stuff that completely avoids the "smell" you're talking about by giving specific guidance on style and perspective (e.g., GPT-5 Thinking can do "literary stream-of-consciousness 1st person teenage perspective" reasonably well and will not sound at all like typical model writing).

breuleux• 11 months ago

The only one I was fairly sure was human was #6, and that was the only one I kinda enjoyed. In any case, as someone who reads a good deal, I agree. I didn't think any of the stories was particularly great (not enough to bother ranking them, beyond favourite) so I don't care all that much about the result.

AI can do much better than what is posted above for flash if given more sophisticated prompting.

How sophisticated, compared to just writing the thing yourself?

unignorant4• 11 months ago

In another reply I gave an example of something you can do: https://news.ycombinator.com/item?id=44937774

I enjoy writing so a system like this would never replace that for me. But for someone who doesn't enjoy writing (or maybe can't generate work that meets their bar in the Ira Glass sense of taste) I think this kind of setup works okay for generating flash even with today's models.

lelanthran2• 11 months ago

I was surprised at the result, and even more surprised when I read that one of the authors who did the test got 4 out of 5 wrong, and rated 2 of the AI stories highly.

Looking at my notes, I got one wrong (story 5, dunno what the "name" was supposed to be, assumed that the "name" is something widely-known in culture that brings about the end times, a something that I didn't know about, and so marked it as Human because of a supposed reference to a shared cultural knowledge), and all the AI written stories I rated at either 1 or two points, with the lowest Human-written story getting 3 and the highest getting 5 (Story 1).

It makes me wonder if we are over-estimating the skill an author has when reading based on their demonstrated skill when writing.

IOW, according to my notes/performance, the AI stories were easy to spot and correlated with low scores anyway, while the author(s), who actually produced high-rated stuff for me, rated my low-rated stuff as high.

keiferski2• 11 months ago

I think if you compared the AI stories to works by “top” authors, the results wouldn’t really be as close. No one is confusing a story by Kafka or Conrad with a ChatGPT one.

Because unfortunately, one reason why readers can’t tell the difference between the AI and human authors is because they don’t have much exposure to the greats. The average person reads something like 2 books a year, and they probably aren't reading Nabokov.

vunderba3• 11 months ago

My personal litmus test that works fairly effectively with these AI generated stories is this - if someone asked you "What was the story about?" - could you reply with anything more substantial than the prompt that was given to generate the story to begin with?

Have a read through the 10 dragon stories where the prompt was "Meeting a dragon" and you'll see what I mean.

https://mark---lawrence.blogspot.com/2023/09/so-is-ai-writin...

bubblyworld2• 11 months ago

That's probably true, but as the author points out, it's still interesting to see where the boundary is at the moment. It's a lot further along than people typically argue imo.

keiferski2• 11 months ago

I don’t really find it that surprising - if you write generic low-quality stories, it’s difficult to differentiate your work from an AI writing generic low-quality stories. People have been selling slop stories on Amazon for a long time before AI tools, so describing them as professional authors is not exactly damning here.

mariusor• 11 months ago

With all due respect but at least Robin Hobb, is one of the greats in her domain.

lithiumii• 11 months ago

I love what he is doing but really hope the voting interface was better. Also I wonder what the results would be if there are AI-assisted stories, but maybe real authors would hate to do that.

famouswaffles4• 11 months ago

I put the basic prompt to 5 medium thinking in the API (because most of the writing gains seems to be tucked away in the reasoning mode and because you can't trust the router for this stuff) and this is what i got. It's not flash fiction, more a short story, but i'm impressed.

https://pastebin.com/huGhbX7u

relevant_stats• 11 months ago

I'm very surprised according to results people struggled with identifying [3] and [4] as AI.

IMO both are simply bad and both contain usual telltales in spades (continuity problems, failed or trite metaphors/analogies, semantic failures, overall feeling of 'wtf is even being attempted here').

I'm not so surprised that people struggled with identifying [1] as human - the confounding factor is that this flash story is unpleasantly written, and it's not easy to realize that its failure modes (eg. trying to cram too much in too short a text) are rather human like. And I'm sure the fact that arguably the hardest to digest and rather bad human story opens the poll might somewhat influence the further analyses.

As others in the poll I failed to identify [5] as AI even though in hindsight the telltales are also there. That's because I rather liked it, and as a result it was harder to be vigilant. I also was very undecided on [8]. Finally I scored 6/8, but I wouldn't say it was easy.

Shame that comparing to the previous contest https://mark---lawrence.blogspot.com/2023/09/so-is-ai-writin... is not straightforward. In that one I scored 9/10 while having very easy time (I didn't even finish reading some of them before making up my mind). I also felt completely excused with my only failure, incorrectly identifying as AI the story written in the style of exhaustingly banal fan fiction. But frankly I found almost all the human stories in the previous edition better then the current ones.

In retrospect ChatGPT4 was a terrible writer. ChatGPT5 seems to be an improvement to the admittedly worrying point. Still not impossible to discover though.

However these are my impressions only and it looks maybe I was lucky and I should not generalize it? According to the website people had serious trouble discerning gpt4 writing also 2 years ago. And I'm rather shocked they did. And that they scored some of those banal AI stories positively.

If it's not luck on my part, then maybe discerning AI writing is a skill very different from 'writing' or 'being deeply interested in literature', skills of people who usually frequent this blog?

spongebobism2• 11 months ago

Maybe this is cope, maybe it's my dislike of modern fantasy, a genre that keeps repeating the same narrative elements to get the same reactions out of the reader, but I found the stories kinda bland, and to me they all seemed like something ChatGPT could have written on a good day. Compare them to Behm-Steinberg's Taylor Swift [0], which has a far more interesting premise, significant character development, and good style (to be fair, story number 6 also has relatively good style). Or to Samatar's The Huntress [1], a poetic piece of flash fiction that is willing to leave much more ambiguity and really shows what's possible in terms of style.

[1] https://tinhouse.com/the-huntress/ [0] https://gulfcoastmag.org/stories/2015-barthelme-prize-winner...

paulmoulinex123• 11 months ago

A lot of folks here seem sure that models cannot hold long-form context. In my experience that is not true. The real issues are model choice and how you frame the task. For the past few months I have been generating full-length fiction on topics I care about, and the quality is good enough that friends in blind reads get just as absorbed as with solid human-written work.

Example of a long book I put together in about 20 minutes: docs.google[dot]com/document/d/1mA-q1ugWRa6BaOghUTH17kYdjHpaNPIWgoNKg6OvOko/edit?usp=sharing Yes, there are rough edges, but the character arcs and the lore stay coherent from start to finish.

Disclosure: I am building an iOS app that lets anyone spin up a book with any cast and starting situation, in a voice close to a favorite author, and then steer the story while reading. On-demand long-form content is already here. It is just not packaged very conveniently yet.

AI vs. Professional Authors Results

Related Stories