Not to be intentionally contrarian, but I have thought a lot about this:
I think good fiction is actually the antithesis of AI, fundamentally. LLMs are really only capable of combining different things into output. This can give you some creative results, but it is ultimately kind of limited. I cannot imagine that prompting an AI a thousand times would give you something truly original in the way many high-quality fictional stories are. The magic and absurdity of life is a necessary thing to create such works.
> LLMs are really only capable of combining different things into output
> The magic and absurdity of life is a necessary thing to create such works.
These two points seem slightly conflicting. Why if a human is truly more creative than AI, must it experience "the magic and absurdity of life" if it can be creative without "combining things into output"? Is a creative human then just not splicing and recombining their experiences into their creative output? And therefore, if an AI had more data that mirrors that experience in some way would it be sufficiently creative to create something "truly original"?
My view is that intelligence and creativity is all just purely recombination. AI can do that well just as humans can. The gap is that the data humans have access to is much more analogue, emotional, and grounded in the real world compared to the data that AI is currently trained on.
> And therefore, if an AI had more data that mirrors that experience in some way would it be sufficiently creative to create something "truly original"?
An important distinction seems to lie in a human having original experiences to draw from, while an AI has only "data that mirrors" experiences.
> while an AI has only "data that mirrors" experiences.
True for a lot of current AI. With more robotic bodies, loads more access to cameras/audio/satellite/science measurements, an AI can have original experience on a level that can only be described as god-like compared to humans.
You could argue AI already partly has access to these measurements. You could easily tap into smart sensor data etc, same voor video/audio etc.
The part that AI can't (yet) truly replicate is 'feelings/emotions'. Experiences such feeling dread, loneliness, happiness. Or being anxious, nervous or having goosebumps. You can describe all these states but they are differently experienced for each individual and also dependent on each unique situation.
Ofcourse this also then comes down to how you exactly define emotions.
We don't know that Claude doesn't feel dread when it looks at my code. To be fair, these emotional circuits are the result of evolving in various environments. AI might well have completely new emotions that we just can't comprehend.
And at this point you’re far off into science fiction, and beyond reasonable discussion of this.
Even then, it’s not really clear to me that an android with the exact appearance of a person will have the same experiences. They aren’t biological creatures with millions of years of evolution behind them.
If it has a subjective viewpoint, i.e. sentience. Otherwise it still doesn’t actually experience any of those things, and can’t tell actually you what it’s like to experience those things. It can give you the data readings though.
Wouldn't it be funny if they missing thing is actually the body, and having one to give the subjective viewpoint something to center on directly leads to emerging sentience?
Though I think there's at least one more important thing missing, namely the ability to form memories - to incorporate recent context into the model's weights. I don't think it's a coincidence that this is a very complex process in humans as well (involving sleep and dreams as neccessary mechanisms).
ChatGPT has had more conversations than you or I have. And it's conversed with people from a wider range of backgrounds in a wider range of situations.
The difference is huge. An LLM conversation can contain within it a batshit crazy model of the world and still produce a plausible stream of tokens. No skin in the game. If you don't test your ideas against reality, you don't have a model of reality; you have a model of written human thought. It's critical to understand the difference.
I think this is actually a tough one. Yes, I agree that an LLM can never experience this. But -- when a real human needs to translate that experience back into words, it can actually do so just fine.
In other words, when a real human needs to translate an experience into words to share it, A LOT of information is lost. The LLM might actually be better-suited for the the task in your example, even if the point you were driving at stands.
I don't think it's all that difficult. Language is not a 1-1 match to reality. Ergo any entity which is inherently language-only, will fail to accurately describe the experience.
Not if the only result you can judge is also language-only, and the entity has access to a large corpus of descriptions produced by people who have had the experience.
Then you're just comparing descriptions, and you're also dependent on the language skills of the experiencers, and you're still stuck with the problem that language is not a complete map of reality. So an LLM may be able to give you the optimal description of a feeling, on par with what a person would give you; but it won't be equivalent to the actual feeling.
>and you're still stuck with the problem that language is not a complete map of reality.
Definitely agreed, but at least when talking about experience we've been stuck in this position for most of human history. How do I know what it's like to eat a hamburger? Perhaps, like the cute visual in Ratatouille there is a deep inner experience but (also like in Ratatouille) when I need to communicate that experience with another person I'm stuck relying on language, and the language is found lacking.
I know it's just a kid's cartoon, but interestingly Ratatouille was making this exact same point -- he really lacks any meaningful way to convey his inner experience. It's locked behind language. The film tackles this by showing an Remy's complex and interesting visual, but his friend only gets the fuzziest, dullest visual when he attempts to communicate it.
Remy's immediate rat & family peers actually _never_ get to understand his experience, and instead later help him out of love an allegiance.
That is descriptive knowledge. Not firsthand qualia.
You may know every detail of the American Civil War, from every book ever written on the topic. You may read every journal from every person that wrote one about it.
You do not have firsthand knowledge of what it was like to participate in the war. Any assumption of that knowledge is a (false) generalization of what you think a person “would have thought or felt” based on contemporary ideas of the personality, psychology, etc. Reading a journal entry about a soldier seeing his comrade killed is not equivalent to actually having that experience yourself.
Stick a camera and a touch sensor on the machine, now it is experiencing the world.
Besides that, there is a lot of value in redigested experience. Quite a lot of history is only really examined in a larger context, where the person describing it cannot have experienced all the original events.
Ah, I see the goal posts shifted. I thought the topic was whether an AI could "give you something truly original in the way many high-quality fictional stories are." Which was then followed by the idea that you cannot have "truly original" outputs unless you also have the corresponding original experiences.
Implication is that virtually none of the novels ever written which include wars are actually original works, since vanishingly few novelists have been in active wars.
That’s a pretty lacking theory of knowledge. How can you know what taste is, if you’re never tasted anything?
If you read a thousand books about living in New York City, is that the same thing as the experiential knowledge that comes from living there in person? I certainly don’t think so.
That's literally what the conversation is about: what is knowledge?
You cannot claim that descriptive knowledge of something is equivalent to experiential knowledge of it, unless you have a theory of knowledge explaining why this is the case.
You asked for a description: "Can you tell me what it feels like to eat a hamburger"
I am sure I Fable (or GPT-5.6-Sol) could answer this question very well. It can probably even tell us what it feels like to eat a hamburger for the first time when you spent the first 18 years of your life as a vegetarian.
I don't think you understand the premise of the question.
Asking another person or Fable what it feels like to eat a hamburger will give you a description of what it feels like. It will not give you the actual feeling. You are merely repeating words that someone else has told you. The description of something is not equivalent to the experience that comes from doing it yourself, as an experiencing subject. Reading a war journal about being in battle is not equivalent to the actual experience of being in battle.
To illustrate the point again: if we gathered ten people that all spoke different languages and gave them a hamburger, they would all gain the knowledge of what it feels like to eat a hamburger, even if none of them are able to communicate this to the others in language.
I understand the premise. My answer was meant to illustrate that, no, the 'original experience' is not needed, and that AI can produce a good answer relying only on '"data that mirrors" experiences'.
"Is a creative human then just not splicing and recombining their experiences into their creative output?"
I do think so. The difference is that, even apart from synthesis, the creative human knows how to filter out the shit ideas from the clever, subtle ones—LLMs seem to just run with the first…um, thought that… um, comes out of their head.
I suspect knowing that you've on to something special probably does require "human experience", woe, melancholy, "the human condition", reflecting on the brevity of life…
> My view is that intelligence and creativity is all just purely recombination. AI can do that well just as humans can. The gap is that the data humans have access to is much more analogue, emotional, and grounded in the real world compared to the data that AI is currently trained on.
Yes!
Aside from the more fundamental questions about the nature of cognition and what constitutes intelligence, there's a practical matter: humans have life to learn from. We have a lifetime of persistent memories, an identity, a social life, physical interactions with the world of many qualitatively different kinds, and we're taken on a lifelong journey into different studies, workplaces, and living environments by the biological need to survive.
Even if human creativity is in some sense also recombinative at its core, we have a very different set of inputs from LLMs. Even under the (almost certainly false) assumption that human condition is basically similar in structure to an LLM, we should expect some differences in outputs based on these profound differences in inputs.
>These two points seem slightly conflicting. Why if a human is truly more creative than AI, must it experience "the magic and absurdity of life" if it can be creative without "combining things into output"?
I agree, yet I also think AI is extremely limited in terms of new creativity.
My guess is that humans operate on some deeper level. Much like the grokking model example where it easily trained on the training data, but only later did it really start to understand the deeper problem space and 'grok' the problem, I think LLMs are doing something similar with human thought. LLMs can throw more compute, but they are still lacking some introspection/drive that leads to the better creativity of humans. Grokking this from the training data might not be possible, or it might take some fundamentally different approach, or even an imperfection in humans that LLMs don't have (ego/drive?).
This isn't to say that LLMs could never do it, but the cost to do so might require magnitudes more scale in training times and size. Or it could be in the next model released. Or maybe they already can, but that gets trained out in favor of overall more general correctness accross all domains.
I think you're just neglecting the fact that LLMs only exist in the virtual world, whereas we're interacting with a world that's richer and has much more to offer. The virtual world is still just a mere subset of our world, but whether we want to perceive it as such is up to each individual. It certainly feels like you're conflating digital with analog, and a digital signal can still only approximate an analog one.
I still doubt that we have the knowledge to describe all of our human desires as mere results of our brain interacting with the nervous system either. There are many processes going on besides analysis and synthesis. If I plan on doing something extraordinary, I don't only sit there, think, and try to rearrange some pieces from memory. I'll go look around, touch, taste, re-explore and expand the perception my environment has to offer. Further, my perception can vary by the minute and I could even try to bring in a touch of chance whenever I feel like.
Wittgenstein's "whereof one cannot speak, thereof one must be silent" might be fitting here, yet language still continues to evolve.
"…LLMs only exist in the virtual world, whereas we're interacting with a world that's richer and has much more to offer."
I have a gamer friend that believes, having played 100's of 1st-person games, that he has had rich experiences because of it.
I reflect back on a day my family and I set out to hike through Arches National Park (still a Monument at the time, I believe) and the heat was such that my oldest daughter seemed in danger of going into runaway heat exhaustion. I began to panic as we were so far along on the hike already that even returning to where we parked seemed to pose danger for her.
There was little to no "coverage" (shade) along the trail and the heat and sun were intense that day—we had passed almost nothing. Finding the smallest of bushes to provide minimal shade for her, I scouted ahead on the trail to look for some kind of larger area where there would be enough shade, perhaps a breeze even, to allow her to bring her core body temperature back down.
Fortunately I found just that. (And it turned out that, in the news later, two hikers within a few hundred miles of us had in fact died that day as it was a record heat wave for the area.)
I think somehow that me and my daughters experiences that day, when actual death was perhaps on the table, was something a "virtual environment" can never really offer.
Is a creative human then just not splicing and recombining their experiences into their creative output?
No, they aren't. I'm not sure why it's such a popular narrative that all art is just some kind of recombination of existing things; it's not. Actual lived experiences are unique and impact what is created. Novels written by people in post-Revolutionary France are not the same as novels written by post-WW1 France. Events in history shape culture and introduce new ideas and experiences that weren't there before.
The gap is that the data humans have access to is much more analogue, emotional, and grounded in the real world compared to the data that AI is currently trained on.
That's the problem with your argument, right here. Emotions are not merely portable "data," and framing this way is a problem from the start.
I don't think you get human experiences without actually being human. The most perfect LLM copy of a mind is just the copy of externalized output.
It's often said that creativity involves reuse because that's clearly part of the process. It's like a warning against another narrative, where a reclusive genius outputs great works made from scratch, without being part of culture. Yet there's some truth to both. The thing is to be inspired by the culture and then be defiantly original and work away stubbornly in private and also talk to your peers and your audience. It reminds me of how we used to play as kids: oh you're doing that thing? I'm gonna do that thing too ... when you see the big reveal, you're gonna be very impressed! I'm doing it differently! Look at how I'm doing your thing that you came up with!
"I'm not sure why it's such a popular narrative that all art is just some kind of recombination of existing things; it's not."
I would like to see art that was "created in a vacuum" so to speak. I am not aware of any. A first time poet who had never read any poetry or prose or heard a tale told? A painter who had been blind since birth and just handed a canvas, oils?
(To be pedantic, the person you quoted said art was recombining their experiences which you took as existing things.)
Agreed! I find AI to be pretty decent at documenting and explaining technical stuff, which is a subgenre of non-fiction.
I once tried telling it to write a fantasy novel for me because I didn't know what to read next. Gave it some examples of books I have enjoyed in the past so it would have an idea of my taste...God was that awful. Absolutely terrible idea.
AI is not coming for fantasy authors jobs any time soon!
Models are increasingly showing their ability to extrapolate into the human unexplored (math proofs being the most apparent). What gives you confidence the absurdity of life is uniquely difficult for models to source?
Mathematical proofs are hard to come to, but easy (or “cheap”) to verify. Also having a math proof processed through a transformer to derive an output that matches existing proof patterns kid of makes sense. The former allows you to generate hundreds or thousands of iteration and cheaply verifying its correctness until you hit the right randomness level. Generating thousands of fiction books then verify them doesn’t really work.
Most good fiction authors are trying to communicate a story that’s forming in their minds. Even when they get stuck, they are usually not stuck because they just don’t know anyone to ask. It would’t be their voice anymore. Some do seek input and those might experiment with llms for ideas. Using llms for research, technical questions, grammar, synonym sentences, word selection, etc makes sense too.
I’d agree with OP for people generating full stories through llms. While I say that, I know a very young person who told a while back (like early ChatGPT 4 time) they use it to write fan fiction for books or characters they like.
Math seems entirely unrelated to the human experience to me, at least in the sense of “this thing happened to me and I will transmute it into a work of art.” Math is pure abstraction.
I will readily admit that I didn't read Dostoevsky, and I don't continue to read Dostoevsky, because I've compared him against a swath of other fiction writers and authoritatively determined with some objective reproducible standard why his writing has been impactful and changed my life. I heard about Dostoevsky from a lot of people talking about depression when I was depressed, he resonated with me. Then when I came back to him half a decade later, having resolved more of my issues, I was able to evoke similar feelings but through a more mature lens.
Do I reflect on the objectively correct structure of the story? Do I "agree" with everything Dostoevsky is saying? Not really.
I don't know if this offends literature people or not. But this sure as hell is incompatible with the AI bro paradigm in which all thought can be objectively reduced to some shared standard to benchmark on.
My position will be read into as deep irrationality from many opposing positions, but it's fine.
Now, none of this doesn't "preclude* the ability for AI to have the same kinds of network effects. Of course it can - that's what I've been saying for the longest time. But don't confuse, becoming integrated with society, with this ideal of superhuman X.
Now of course, you could argue, "there is no such thing as anything else besides becoming integrated with society. There is no other standard that works." Very good! Darwinian-adjacnet positions are interesting. This is a position - but then don't claim it's equivalent, or god forbid, the exact same, as the "super AGI" ontology. These require fundamentally different assumptions about what justified what, and an attempt to try and pretend otherwise is sophistry.
I have also thought about this a lot but have less developed views on fiction (short version: I don't currently want to read an LLM-written novel, but I expect that will change in the future and there will be some I enjoy)
Not every book fits this pattern of course. But even a more straight forward book (https://www.goodreads.com/series/139929-designers-dragons) will often rely on sources that are not easily accessed by an LLM, they may not be digitized, or they may not be accessible in your common digital collections.
I 100% think LLMs can do good synthesis work, don't really think we have the tooling/checking in place for them to write a competent non-fiction book yet, but that will certainly happen, but I still think the best and most interesting non-fiction remains firmly in the realm of humans.
(Also, a lot of non-fiction is memoirs, travelogues, etc. which LLM can't really replace unless you're dictating to it.)
While you are right that humans can probably find some original ideas that are not just combinations of inputs, the universe of things that humans will find interesting to read is not that vast, and that’s why there are only really like 7 types of stories, and those are the ones that survive through a process of natural selection.
Everything else is just not that interesting to a human.
I would say the inherently chaotic and continuous nature of the human brain can create novelty in a fundamentally different way than LLMs can.
Like a human can reach into the chaos of their own emotions and pull something unique out. There is some layer of the human experience where ideas can ferment and change based on all your memories is weird ways, and LLMs completely lack this, which makes things LLMs create very blunt and obvious. Like there is no nuance to anything they do.
Well since human culture has developed from very limited stone age culture into what we have today, we must conclude that at least _some_ people are capable of generating something that is genuinely new.
So now we'll have to investigate what "original" means?
From what I can see, AI is absolutely able to create new work. Things that are not in the dataset, but look like they could be. This is pretty much what generalising means, they can make a wide variety of things that aren't actually in the original set but can hide in that set pretty well.
When you are an original piece of culture, what do you mean by it?
Can an AI write a story about living as an artist in NYC in 2026, the cultural environment, etc. without that information being added to the dataset immediately as it happens?
It seems pretty obvious to me that there are cultural trends which are happening instantaneously and aren’t merely some sort of consequence of past historical data.
Ideally the AI would get none of its world knowledge from its model weights and would rely entirely on context. This is just an ideal though, we haven’t figured out how to completely separate world knowledge from the model.
But how could something be a cultural trend without some sort of anchoring in history?
I don't see why a computer (or a person) couldn't make up a plausible sounding story about a NYC artist of the present day? You would just reference things that don't change that much (artists having a boheme type existence, wild inequality in who can live off their art, NYC being a sort of place where people go to become successful).
> LLMs are really only capable of combining different things into output
In terms of how writers think about creativity this doesn’t reflect reality- there’s the old writer’s saying that there are only 7 stories (and other variations of this idea) and that writing is about creatively remixing these well-worn ideas.
The LLM’s capability to write acceptable fiction and nonfiction is coming soon if it’s not already here. (I think right now it still needs some high level input from humans, depending on length and topic)
It’s clear to me that, especially as LLMs get better and better, there’s only one real difference- it’s that I don’t want to hear what an LLM has to say. I’m only interested in what other humans have to say. I feel the same way about AI music- even if it sounds ok, even as good as the kind of unoriginal pop music that’s not AI made (but is a kind of it’s own slop) or a bad committee made hollywood movie, it’s still fundamentally more interesting than something that’s been generated. Even if it’s 100% fiction it’s still based on a real human’s life experiences.
This is a bit like saying “there are only five fundamental flavors” and everything is some combination of them.
Maybe, but that has basically zero effect on the experience of actually eating food. Just because the form may be the same doesn’t mean an LLM will be able to supply original content.
I’m not sure I understand the analogy. You could say that the reason why artificial flavors work in food is because we know how to recreate flavors from first principle chemistry. The dislike of the artificialness of cheetos also doesn’t stop many many people from eating them.
I would say human made slop writing isn’t any better or worse quality than LLM writing.
Knowing if it was made by a real person makes a difference to me, but I also know it won’t to everyone.
Just because the form of a story is the same as another story does not imply that the content is interchangeable.
I could write you a story right now about being an expat living in Poland in 2026. It’s unclear to me how an LLM would have the cultural knowledge or experience to write such a story that is actually accurate and up to date, without already having something like my story already in its dataset.
Yeah but you would just be doing the same thing: mashing up some known information about humans and Poland. You would pick plausible names like Marek or Piotr, plausible places like Warsaw or Gdansk, and some random flavour of story that is either completely general across cultures (hero's journey) or some local tale from Poland.
Shakespeare did the same, so I wouldn't blame you for it.
I am an expat in Poland, that’s why I used it as an example. The point being that I have knowledge of that experience that is not written down in an LLM dataset.
Superhero Wars Trek 42 - the latest franchise installment, billion dollar Hollywood budget, financed and produced by billionaires, audience tested committee-made script, lots of people involved, designed primarily to push merchandise.
Personal Story - a movie about someone's personal reality or fantasy, something typically not made by studios because the story isn't profitable to a wide audience, painstakingly generated and edited by one person using AI, designed primarily to tell a story.
I put up a different strawman because your strawman is too beatable. As humans, of course we're all interested in what other humans have to say. But you haven't explained why you feel people can't use AI to say things, or why AI output can't be based on a real human's life experiences.
Personally, it seems like movie studios are the ones with only 7 stories to tell, and I've seen the movie and its sequel, the prequel, the requel, the remake, the gritty reboot, the musical, and the animated series. I'm excited that AI will let people skip the studio gatekeepers to tell some new stories for a change.
Exactly. Shows us what is true in the sense of the experience of being a human. It is resilient against slop because good literature has experiential depth. Maybe LLMs learn to emulate this some day, but I doubt it. Not enough money in getting it right.
I do think book prizes are a better-than-average signal but I previously volunteered with a book award. I will caution that pretty much every publisher mass-submits these books for consideration in every remotely relevant prize. It is a cost of doing business (similar to how photographers pay to enter photography competitions to try and win the "award winning photographer" title, or businesses submit dossiers with consideration fees on why they're one of Michigan's top 100 places to work).
There are often so many books and so few willing qualified readers that which books get an award can either be completely arbitrary, or comically easy.
For example, the NCR Book Award in your book corpus faced a big scandal when it was revealed that the judges did not read the books themselves. [1] The PROSE award is so comically large that an ordinary category finalist or win is usually overstated in prestige and value.
Awards are also a product of their time, representing biases among the judges, literary preferences from the readership, and straight up nepotism/favoritisim/elitism. Look at the Hugo/Nebula/Locus and you'll find a lot of so-so winning work throughout the years; especially when you look at the other nominees.
People’s tastes in beer and wine vary so dramatically (and there are so many styles) that it’s not clear to me that you can actually have a “best of.”
The beer and wine I enjoy the best my wife finds completely unpalatable. For wine, I enjoy a dry Grenache; Mas de Gourgonnier is a good one. For beer, I enjoy a beer with body, a malty profile, and balanced hops; there are many mass-produced styles that fit this, but I like Sam Smith Organic Lager, and Leffe Brune. I’ve also brewed many beers myself and one of my all-time favorites was an all-grain formulation of a honey ale [1].
Such is the nature of beer and wine (and HN) that someone is going to “disagree” with me. That’s fine. Just go try some things and see what you like.
I would bet that how your brain stores information that you read from long-form text is very different from how it stores information you acquire from chatting with an LLM. When I read something challenging or new to me I spend a lot of time thinking about how what I'm reading matches my own experiences or knowledge. Although I'm a fairly fast reader, it often takes me a long time to get through difficult pages since I have to stop and think about what I'm reading. I seem to be doing a lot of integrating and reorganizing my thoughts. When interacting with LLMs it feels a lot more like I'm just receiving knowledge passively and I don't think it gets integrated as well. Not sure why this is and its somewhat counterintuitive since I don't think I'd have the same experience with a human tutor.
> When interacting with LLMs it feels a lot more like I'm just receiving knowledge passively and I don't think it gets integrated as well.
This sounds like it's more down to how the individual uses the tool. I am not someone who has even been particularly good at reading > learning the thing. I've only ever been capable of learning by doing and with capability to interrogate on the points I don't get. LLMs allow that in a way that is just not feasible with any human being whose tolerance of me would diminish rapidly.
In most cases, I will be writing down my understanding as I would in isolation from a primary source, building flash cards and actively practicing what I have learned, only with more capability to interrogate on the points I have difficulty understanding or need clarity on. Effectively, I am doing the following in a capacity I personally never had via any other means:
> I seem to be doing a lot of integrating and reorganizing my thoughts.
If you are using it as a slot machine of knowledge, and going from receiving > doing with no intermediate step I see how outcomes could differ.
Seconded. At a minimum, you can literally tell the agent to guide you but not point you and only help when you are stuck if you're actively trying to learn something. It's been a help for me to bridge a couple of gaps I'd struggled to cross before (mainly hardware things).
> This sounds like it's more down to how the individual uses the tool.
I genuinely agree with a lot of your comment but lines like this drive me crazy. Anytime somebody has a critique of AI, so many people hand wave away with “well you’re doing it wrong.” If we can just chalk up every negative thing with AI to “incorrect tool usage” then we’ll never honestly assess it.
I didn’t say they were using AI “wrong.” That’s a stronger claim than the one I made.
Their assessment was:
> When interacting with LLMs it feels a lot more like I'm just receiving knowledge passively and I don't think it gets integrated as well.
My point is that the passivity is not necessarily an inherent property of the tool. It depends heavily on an individual chooses to engage with the output.
Saying LLMs make learning passive is a bit like someone else saying books make learning passive because you can skim them without thinking. When we all know instead of reading a book passively you could in addition annotate it, question its claims, summarise it, test yourself, and ultimately apply what you learned. The medium permits both. The same is true of the output of an LLM. You'll get out what you put in and nothing more.
You can use it like a slot machine, repeatedly pulling the lever for another answer without reflecting on any of them. But the fact that a tool permits shallow engagement does not mean shallow engagement is the only option available to you.
That also doesn’t mean every criticism of AI should be dismissed as “incorrect tool usage.” AI has real limitations and deserves serious criticism, and I assure you, I am not popular with the AI hype bros for this reason. However we still need to distinguish between limitations of the tool and consequences of how someone uses it. Not every potentially negative experience is rooted entirely in the technology itself. Sometimes user behaviour is a significant part of the cause.
I listen to audiobooks while I walk and hike, and sometimes when I recall some particular thing I've learned, I can also recall where I was hiking when I heard it.
I think there's a lot to learn on how we really process information.
I once heard a lecture about „learning how to learn“. There are already a lot of well-researched things one can exploit about the way we process information. What you describe is known as retrieval cue. Our minds seem to be associative^1. It also seems a lot of the learning/processing happens after the actual learning process, which is why resting is important. Spaced repitition is a further method well-proven in practice.
^1 In the early days of AI, the concept of Hebbian learning was popular („neurons that fire together, wire together“). Its implementation is even simpler than gradient descent used today, but never caught on https://en.wikipedia.org/wiki/Hebbian_theory
Perhaps not the same one, but there's a fantastic course by Barbara Oakley by this name. I viewed it quite a while back, but the modern incarnation seems to be [0]. In a similar vein, the book "How we learn" by Stanislas Dehane is also very good!
Same here. I always marvel at this. These recollections can even be years later, and the memory of the place in which I heard the remembered thing is vivid.
There's a podcast I listened to as it came out when I was a teenager. I'm now relistening to it over a decade later, and I distinctly remember walking home from the bus stop with it coming through wired earbuds from my ipod.
That's spatial memory! I have long walk-and-talks with friends, and sometimes I forget minor details of previous discussions, but not where we were when it was mentioned.
A lot of my friends can't do same trick. But the fascinating part is that it still works for them! Usually they can recall whatever I forgotten after I described the place we were in.
"when events are represented in memory, contextual information is stored along with memory targets; the context can therefore cue memories containing that contextual information"
https://en.wikipedia.org/wiki/Context-dependent_memory
What you're describing is very close to what education theory calls "assimilation" and "accommodation". When we assimilate knowledge, it just fits into our schema of understanding. When we accommodate knowledge, we need to change our schema, and that is where the feeling of being challenged (and often the feeling of profundity) comes from.
For example, a child learns that "foreigner" means someone from outside their country. Then, when they're 11, they go on their first holiday abroad and realise "Wait! _I_ am a foreigner here!"
So, maybe one way to frame what you're saying, is that LLM output tends towards being easily assimilable.
In college I used mathematica a lot while taking linear algebra. I ended up having to relearn a lot of that math later since I never really understood it at a deep level, although I was able to use it and apply it under class conditions. I feel like a lot of my LLM derived knowledge is similarly superficial.
That's absolutely right. Math is difficult to understand without writing it down on paper. As one of my professors said, "something magical happens when you write down a problem from scratch and work it out". LLMs make people believe they are improving their productivity and that may be so, but it also removes that magic and removes that deep learning.
Any such bet would be grossly misguided given how different people acquire information differently. Just because you don't like using LLMs doesn't automatically extrapolate to a general realisation about the human brain or how one approaches knowledge.
Your subjective experience matters in this case. Almost completely.
I like using LLMs and one of the main things I'm working on would be impossible for me to tackle without them, but I do think they cause some skills to atrophy despite being a big productivity booster overall.
Thank you for this. I sorted the "technology" and "science" sections and saw a few excellent books that I've read and a few that I would really like to read. This motivates me to start setting aside "reading time" again every day, since I've lost that habit.
The recent "society & culture" books gave me some good book club ideas.
Bug report: filtering by "award" appears to be broken for some awards. If I select Pulitzer or National Book Award, no books show up, but I can find books with these awards by browsing.
Awesome, that's what I was going for. I originally made this for my own use and got kind of obsessed with building it out once I started finding unfamiliar books I enjoyed using the semantic search and category browsing. Glad to see it out in the world hopefully doing the same thing!
One thing that would make this information even more useful and searchable would be a JSON dump of your dataset. You could serve it statically on S3 or similar, to keep hosting costs to a minimum.
Of course, having created a cool thing generates no obligation for further work on your part! I'm grateful that you did this work and made it available for free.
Did you build your semantic search index based on the full text of the books, or just the reviews and descriptions?
Already tried to do this in a basic form. Advice from experts would be welcome as it's my first time trying to make data and an API available: https://book-prize-index.vercel.app/data
Also I just had to upgrade my Vercel plan due to traffic, which is very welcome and appreciated, but if anyone wants to donate, that would be really helpful! There's a Stripe link at the site: https://book-prize-index.vercel.app
I recently did this! I fixed my phone habit and my TV habit, and I just read no, the family has silent reading parties instead of movie night.
Here’s what I did:
1. Install Jomo on the phone. There are other apps, but basically it makes me wait 5s before I can open a brainrot app, and then I can only use it for 5min, and every time I do this I have to wait an additional 5s. It resets at midnight. This introduces the necessary friction.
2. Paper books. The house is now full of paper books. Book seems interesting? Buy it. Books everywhere. Real books.
3. The actual habit. I stack habits. Right now I have a daily workout routine. First thing I leave the house and either run or go to the gym. I bring my book, and afterwards go to the cafe across the street for a coffee. At night, I read before going to bed. Once you finish a few books this way, you’re in. TV now is actually hard to pay attention to, and I feel kind of… dirty doomscrolling now (like eating junk food, it tastes good but you feel gross during the experience).
I bought a tiny e-reader [0] and carry that as well as my phone. When the temptation to doomscroll triggers, I pull out the e-reader instead and read a book.
I tried attaching the book to the phone as their marketing suggests, but that didn't click somehow. Having both out means I would check the phone and then not look at the book.
Working so far, but it's only been a couple of weeks, so let's see.
Yea, for me my Kindle just didnt work. I always have had one since they were first released, but it was always hard to keep the habit between book.
Having a nice library of paper books, for whatever reason, has worked a lot better for me. I think it’s passively seeing them around the house and in the background always considering what is next.
This last round I couldn’t decide what to dig in to after I’d finished True Grit, so I spent a few minutes browsing what we had on hand and it was just… a very pleasant experience.
Also, unless you pirate ebooks or have access to a library that offers them, used paperbacks are actually much cheaper than buying ebooks.
Yeah I pirate my ebooks unless they can be bought DRM-free (which a few can). On principle, because I want to actually own the things I buy.
I used to have paper books. But I found they were just a massive pain in the arse to carry around, and move. They became very expensive, very inconvenient wallpaper. I gave them all away to a charity shop and bought an e-reader instead. Now I carry around 250 books in my pocket wherever I go. It's liberating.
Thanks for sharing jomo. My hack has been to read on my kindle while standing or even walking short repetitive steps (in surroundings or patterns where I can't hurt myself). This has been a game changer while waiting for kids soccer games to get done, lines in the mall, baggage claim etc. It also helps me avoid sedentary reading while sitting.
Suggested additions: the Axiom Business Book Awards, Library Journal Best Books of the Year and Booklist Magazine's Editor's Choice Awards. There's also a few popular substacks that have big sideshows in regular book reviews, but I wouldn't even recommend my favourite for a public aggregator like this.
Original author/creator of the site here - thank you! Will add the Axiom one. I am on the fence about whether/how to add "book of the year" type lists as they are somewhat distinct from book awards, but I do think that would probably be the next step to get more books in the corpus.
And any other ideas that HN readers have for awards to add would be welcome. Currently it's probably too history-slanted since I'm a historian and knew those awards better.
It may be hard to parse the lists but if they're still available somewhere online under the current administration the past annual State Department, CIA bookshelf recommendation, Army Chief of Staff, Navy CNO and Marine Commandant reading lists may have some valuable additions. To my knowledge there aren't any analogues to them in the rest of the Anglosphere but some LLMs will probably point out ones I haven't thought of.
I can point out that the UN agencies' respective reading lists for their staffs' professional development are 99% internal UN papers of very limited interest to external audiences but I haven't had a professional reason to look at any EU agencies to confirm or deny value.
Additionally, the Financial Times has a second set of book awards along their main awards which you've not listed, the FT reader's best books list.
The article resonates, there is great serendipity in somewhat random exploration of books. But then the web app seems to be the antithesis of this? I clicked a few categories and the top ranked books all had the “popularity contest” smell, sort of the antithesis of random exploration.
The history section of the book price index is incredibly america-biased. History of america, american natives, japan after ww2 and vietnam in the last century.
And the irony of it is that the U.S. is actually quite history-poor, when compared to Europe, North Africa, Middle-East, Persia, China, Central Asia and India.
Can I make a feature request? I would love to be able to search by author. I recently started Caro's LBJ series (it's amazing so far) and I'm curious how many awards he's won. Anywho, I love non-fiction so I plan to visit this website often. Thank you and great work!
More generally, just sort by any of the column headers (year, title, publisher, author, score, ...) by clicking on it. In some views it works, in some it doesn't. A bit unintuitive.
The toggle between "traditional" text and "semantic" search is cool and new, but maybe the toggle could stay visible (it's there on the home tab, but then not anymore on the books tab).
At least one of the links on the post is still to localhost:3000 instead of the vercel page. I don't want to make a substack account to tell the author so hopefully this information finds its way over there.
Thanks for this!
I've been looking for a good book on the Spanish Civil War, in either Spanish or English. The one American book I was recommended by someone, I didn't like. I'd like to keep searching regardless of the book's language or what country it has won prizes in. Could you also include book awards from other non-english speaking countries?
The war by itself is a far less interesting affair than how it happened, and what made the country be OK with fighting. For that you need to start at the end of the previous dictatorship, and cover the entire 2nd republic. Then the war starts making sense.
So I'd read The Collapse of the Spanish Republic, by Stanley Payne. He was a journalist in Spain through the entire period, so most of what he does is provide a first hand account. He never claims to try to be impartial, but he's pretty nouanced, as he knew a lot of the players personally.
I've only read Homage to Catalonia out of those, which is a personal retelling from Orwell, are the other books in the same style? What's the difference you'd say between the ones you list?
Paul Preston's "The Spanish Civil War" would be the poster child of the "quality non-fiction" category specified in this post's title. He is, in my opinion, the leading Hispanist historian [1], and I hold his work in the highest esteem. I can also very much recommend his book "A People Betrayed".
Regarding the other two authors in my previous comment: Hugh Thomas was also a leading Hispanist historian [2], and Antony Beevor is a prominent military historian [3]. I expect both books to be in the same vein as Preston's.
(Quick aside for the astute reader that noticed all of the recommended authors being non-Spanish: I find Spanish politics extremely polarized; the Spanish Civil War being perhaps the most polarizing topic. As such, I find outsider's perspectives to be more balanced and objective.)
> Quick aside for the astute reader that noticed all of the recommended authors being non-Spanish: I find Spanish politics extremely polarized; the Spanish Civil War being perhaps the most polarizing topic. As such, I find outsider's perspectives to be more balanced and objective
Makes sense to me, even as someone who live in Spain. It's a hard subject to talk to people about, especially because it never really was addressed as a country, everyone (politicians) just agreed to pretend it never happened, sweep a bunch of stuff under the rug and hope for the best. Then everyone act surprised when things start to bubble again...
I think the approach makes sense, also why you probably don't want to read about US politics from people actively inside of US politics, and it's similarly polarized today. Seemingly, at some point, people just lose all head and reason, and just start backing into their side, regardless of what that means.
I very much have had the same experience as the author with research libraries. I have learned so many life-changing techniques and derivations from books that were two or three down from the book that I was looking for. When I became a research associate, I would often have my librarians ship the books to the branch closest to my office, and I stopped browsing the shelves myself. It wasn't until I went to the main library that I checked out a book several over from the one I had found in the catalog that I had a major breakthrough. Now that same library has collapsible shelving that makes browsing feel like a chore, and it makes me think we lost something valuable.
That said, I had high hopes for the catalog that the author had put together, but for my technical fields, I found it lacking. Trying to find books that would be close to van Soest's Nutritional Ecology of the Ruminant or van der Werf's Animal Breeding: Use of New Technologies, yields a whole bunch of pop-sci titles. The joy of finding these titles at the library in the past was how remarkably understandable and engaging they were while remaining scientifically rigorous. I immediately feel that the pop-sci titles that I was recommended are going to try to sell me something or some idea, rather than letting me explore ideas myself. I guess that means I need to go back to browsing library shelves myself.
On a related note, I was just noting to my co-founder, as we struggle to write good case studies for our website, that I find LLMs are astoundingly bad at writing good prose.
We all know the "AI-tics" that give away a sloppily AI-written piece, but even if you steer them, they still struggle to write consistently high-quality prose.
Somehow I feel that the work of a good copywriter has never been more noticeable.
I recently realized this as well and I think what I’ve discovered is that AI just produces mediocre content in all realms, but you don’t really notice it except in the realms where you have real expertise. With a lot of harness and prompting you can have it pump out something that’s pretty good but by default the next best token rarely produces anything of quality it seems like and if you think it does, perhaps you may want to recheck your assumptions on your expertise of the topic at hand
I do think this explains much of it. But it doesn’t explain why Claude often writes in the clipped tone of that gnomic in-house tech evangelist who was inexplicably hired because he impressed the CEO.
(I may be projecting real life experience onto the LLM.)
I am genuinely fascinated as to how Claude acquired its utterly aggravating way of writing. It’s so much more irritating than ChatGPT, which is already not good.
Yeah the latest batch is really keyed on certain phrases - well beyond “you’re absolutely right!” Of the past. And the readability is trash almost always by default. Probably because they are forcing code so hard that it’s aligning prose into functional groups or something.
Interesting. So you think it’s somewhat emergent, in that sense, rather than a sort of designed tone of voice/writing style specified by supervised fine tuning (or the system prompt)?
I think it's partly because of where they come from to the problem.
I have education/experience in both literature and coding, I have a pragmatic starting point when approaching text while also being able to recognize stylistic oddities, so I get to be the guy editing out AIsms sometimes.
But I've helped other people copywrite where their environment was all org-speak and academic writing, and AIsms don't really stand out in that case. AI is effectively "doing the right thing" writing the way it does for those tasks. Even tho the right thing is often a bad thing.
AI adopted humanisms, we just weren’t used to seeing them at the same scale we do today.
Diversity of writing styles was part of that, but I’d point to vernacular exposure as the larger component. We’re going to go through a period where we try to adapt to a form of “Universal English” for those of us who read primarily English writing.
Other languages’ readers may be experiencing the same dissonance when they come across AI-generated prose in their native language (but I’ll let others validate /reject my hypothesis).
It's like people learning a new language according to the book. Those people sound foreign to a native speaker. If AI was trained on school books for grammar, then they too will sound odd to a native speaker even if their output is technically fine or technically more proper. I known nothing of LLM training and how weighting is applied to educational content vs other sources, but it feels like they were weighted away from modern colloquial speak and towards book grammar. The whole thing reminds me of school literature classes where the teachers comments always felt like I was being guided to a more strained sound and less natural. But what do I know. Lit was my least favorite subject which is well evidenced by my grades compared to my math/science scores.
> we just weren’t used to seeing them at the same scale we do today.
I think there's also a lot of recency bias in it. The same thing that makes you suddenly notice how many people are driving the same model car you just looked at, or how many ads there are for Turbo Encabulators after you read an article about them. A lot of the "tells" people picked out in early AI are tells because they're also really common in the material that the AI was trained on and the styles it was made to emulate. But until everyone wanted "one quick trick" to pick out AI writings, people didn't have any particular reason to need to notice those tells and so they slipped under the radar.
This is right. AI is using human speech, but just doing so in a consistently peculiar way. The em-dash in particular is frustrating, because it's all over high-quality, pre-2022 academic work. But now instead of proper and erudite it's seen as AI-slop. Well, maybe if they didn't use it — all — time — ! It's even an auto-replacement in Word and can be (by one's choice) in LibreOffice, replacing three dashes (and two is replaced by an en-dash).
But when I submit a novel with em-dashes, will sloppy agents and sloppy editors be able to tell that em-dash was deliberately put there by me?
you're missing the point: the prevelance of AIsm language is spreading to people; I'm well aware it came from people.
All these HN posts with "stop with the LLM generated garbage" will slowly fade away because both people and LLMs will talk in a similar format and will be inured to the the distinguishing feature.
Certainly "humans" will keep carving out distinguishing characteristics, but just like "corporate speak" is a thing, AIsm will be a thing.
Random observation: Google's Gemma 4 models write so much nicer prose than ChatGPT or Claude.
Though this might be me as a British reader, simply preferring a rather less American turn of phrase.
I reckon the more transatlantic, english-as-international language DeepMind team have had a subliminal (or maybe deliberate) impact on the way it chooses to write.
Or perhaps small open weights models simply aren't under the same commercial pressure to be engaging and sycophantic and are therefore less likely to adopt the samey overly casual, upbeat, Californian sales assistant manner. (Don't get me wrong, I like this from real human Californians just fine!)
Either way, the default tone is much less showy. I would be interested to find out if you agree.
I am very much an LLM cynic. I am engaging because I must, and trying to learn fundamentals, but I would not say I am overly excited by any of this, just glad that small open weights models exist as a counterpoint.
I loathe the way ChatGPT writes, and the Claude-isms that are everywhere; it is actually quite enraging, especially when you start seeing it in internet comments from people who used to try to write out their own thoughts.
But in my experiments with open weights models I have found I am much less aggravated by summaries and outlines written by Gemma 4, so much that I am happy enough to read them, because they have fewer irritants that take me out of the reading flow.
Though this evening it told me very kindly that my photography is a bit "safe". How very dare it… understand me that well.
I gave Claude Fable $25 in Pangram API credits and, after hundreds of attempts, it was unable to produce a single readable original piece of writing that was not immediately identified as AI.
This seems to be a hard problem for LLMs, as passing would probably require good self-perception ("oh no, I am writing like an AI!") and fine-grained control over its own output ("let's write like a human instead!").
I wonder if it partially because "write like a human" is kind of a vacuous request. Like, it's the objective everyone including me has been saying that we want, but there's no one way to write like a human and and humans don't even have a good definition past "I know it when I see it."
There's a lot of work in the humanities about different aspects of good writing, but that's not quite the same thing. And anyway they tend to assume a pre-existing level of writing ability. Students are supposed to learn good writing through practice; there are rules and exercises but they're incomplete.
I think as much it is that people write by grappling for the right phrase to represent some inner feeling or concept, writing in part for themselves, whereas LLMs write always and only for an audience.
It’s much easier to understand this once you think about other generative forms. MidJourney never just sits down and draws for fun, so fun never informs its art (only the outward appearance of others’ fun, separate from the fun itself). Suno doesn’t waste hours trying to find riffs on a guitar, so its output is never informed by the direct joy of getting it right. Its music is never optimised for playability on a particular guitar with a scratchy seventh fret and a too-high action. Neither Midjourney nor Suno have evolved their styles due to short-sightedness or carpal tunnel.
If you had a human writer who over a long career only ever wrote articles from an outline given to them by someone else, and you had all the outlines and all the resulting articles from those outlines, and you could train an LLM to generate an article from an outline, it still would not be kicking itself frustrated by an inelegant phrase in a prior article, it would not avoid certain phrases out of a passive aggressive reaction to some editor’s note, it would not ever just rush an article because everyone is gathering at the pub, and it would not choose an analogy just to rub the author of a bitchy critical letter to the editor the wrong way. An LLM could not “subtweet”. It could not write a series of articles hoping one important person will spot that they are auditioning for a job.
Creators have unseen, undocumented influences and motivations that inform their work over a long period. I don’t mean to say that these individual influences can be reliably detected in individual pieces of work. I do mean to say that I think their broad absence tends to be felt in LLM writing. As readers we develop an affinity for writers as much as for their writing, and we do this in part because we deduce things about them.
Very elegantly put! So much so in fact, that I finally put down my phone to grab my laptop in order to write this. I have had a similar much cruder version of this thought: LLMs are not human, they don't know what it's like to stub their toe, despite having read probably millions descriptions of it. Even amongst humans there are experiences which are impossible to share, despite us having evolved to communicate quite effectively. Writing is just a small part of it, and cannot replace the actual experience of being human.
Each person writes in a different personal way, so writing “like a human” would actually require a model being able to purposefully make the specific choices that an individual human writer does.
However, general purpose LLMs like Fable have been trained on huge amounts of all kinds of data, and therefore find it exceedingly hard to break out of the grooves carved by that data. They can’t avoid defaulting to centroids and averages, even when they are trying not to. This makes it possible for classifiers like Pangram to discriminate their writing.
A plausible way to work around this limitation would be to train a LLM on a limited and cohesive subset of writing materials, so it would absorb their specific writing style.
One example might be Talkie, a LLM trained on pre-1930’s English text. Talkie is a far smaller and less powerful model than Fable.
And yet, Talkie’s writing is so distinctive that it is often classified as human by Pangram.
A lot of what would be the top "reference" works aren't even that good either, they were a successful marketing phenomenon or had cultural or social relevance at their time. So, you can get a lot of bad prose going by a number of somewhat logical, externally measurable parameters.
More noticeable to me is the lack of the work of a good copy editor which, sadly, we haven't had for a really long time. At least, not on the interwebs. Even the news sites reduced where their print copies were known for rigorous editing saw obvious issues with the various corporate overlords doing serious headcount reductions. The rush to be first to publish reduced even further the time any editors might have had, and then the wide spread use of CMS style articles that slammed output together with something as unintelligent as 'cat segmentFromAuthor1 segmentFromAuthor2 segmentFromAuthor3 > article' where you can tell where each segment started over again with the same basic information as if it was content meant to stand on its own.
Of course, the amount of self published work has also helped make the lack of a good copy editor noticeable. I can excuse self published blogs though. But the stuff released "professionally" has really become farcical.
Worth stating that this is one of the success stories from AI. Someone who has domain expertise outside of programming is able to create a really useful piece of software because the barrier to entry has been significantly lowered. Really nice.
If you are interested in a non-fiction book, check out Veritas: Truth Across Cultures[1]. It's a collection of quotes/saying/proverbs that are similar between cultures, which I call truths. I am in interested in long term thinking, things that stood the test of time. This was primarily put together for myself as a lens to view the world.
Very neat site and write-up, but I found this part amusing:
> There is really nothing “AI” about this aside from the tool that collected the data and coded it, and, crucially, semantic search […]
So really, everything about it is AI. And that’s not a bad thing! It’s okay to simultaneously preach the superiority of award winning books over AI-generated garbage while also acknowledging the same AI as a valuable tool for other uses.
Key word being tool. People get lost in a fantasy that the AI was responsible for decisions and not the human who chose to use it as a replacement for effort rather than a tool to enhance effort, then apply their bias to all AI use.
so much of AI is like, to me, used as was to generate basically dynamic applicaions. if we ever got anywhere with semantics, objects, etc, pushed html further, maybe we'd naturally have web clients that could take objects, natively search, collate, display, put on a map, graph, etc... but AI is like a text only shortcut to the end result we want.
I have a stupid AI app I use now for some bookkeeping which was previously, just an excel sheet. except it has a UI, it validates inputs, and does a bunch of other housekeeping stuff that makes it nice. Did the excel work? yes totally.
Penguin Random House seems to be dominating the leaderboard. What's the story here? Are there only very few non-fiction publishers capable of publishing high-quality work, or is Penguin Random House exceptionally good at it?
Fascinated that the author shares an experience of discovery in libraries. The internet, in its early days felt like that. Of course, no longer.
Libraries are certainly declining in their traditional form. I find it odd that everyone has a digital resources in their pocket yet libraries are squeezing out physical books to make way for more and more computers. Try to find paper copy of the Sony founder's book.... that will be £50 on amazon. Prohibitive. No library within a 10 mile radius has a copy.
I remember joyfully discovering Tony Royce's book in my library a few years ago. That enlightenment will never happen now. Primary knowledge is being lost, churned crude will forever lubricate the delusions of those who have no facility to collate the basis of our understanding.
> The internet, in its early days felt like that. Of course, no longer.
back in the reasonably early days of the web (I would guess ~2000) I stumbled across a webpage that said "I used to collect random interesting snippets in a shoebox, here is the internet version of it". I was delighted enough that I actually wrote to the author to compliment them on helping keep the web interesting, and though I haven't thought about it in ages it clearly made enough of an impression on me that I remembered the author's name. and it's still up! https://www-users.york.ac.uk/~ss44/cyc/index.htm
AIs are fantastic for discovery. Just today I asked Chatgpt to find blogs 1) about old school newspaper cartoons 2) that have been writing for more than 10 years and 3) are run by one or two passionate people rather than a team. Within seconds Chatgpt found 7 candidates. Within another handful of seconds it gave me RSS links to the ones I wanted.
Sometimes. I think this is partly because search engines are so poor nowadays. I have asked AI for certain websites that I know exist but it can't find them.
This hits the nail on the head regarding the real utility of AI right now. The magic isn’t in using LLMs to generate endless reams of synthetic text; it’s in using embeddings and semantic search to navigate high-signal, human-curated data.
My bona-fides here amount to little more than being a big time reader of non-fiction. But IMO, there's a nice synthesis here. Rather than going long rounds of asking LLMs about a subject, I usually end up asking it for book recommendations. It's much better than a google search and you can push it into some deep corners if you go past the surface level recs.
You can get quite specific. You can find texts you wouldn't discover unless you spent years studying the topic. Often these are completely approachable and give interesting perspectives they just get buried behind a wall of syllabi and listicles.
This is how I ended up reading Thompson's 'The Making of the English Working Class' and Graves' 'Goodbye to all That' among others.
Is it even possible to find one these days? I would love to see some real honest amazon kindle stats on the amount of ebooks added over the last 3 years. Even a honest pie chart to show people its not worth it to get them to stop posting them.
"To avoid the tedium, I decided that I would also flip to a random page of every book I shelved and read a random sentence from it."
Also, this is about the lost joy of "browsing".
I wonder how much serendipity played when wandering through a library (or other things in life—even window shopping). But now, everything on the internet, you more or less have to know what it is you're looking for—are rarely "surprised or delighted".
This is wonderful. 30 seconds into trying the site and I’ve already found a couple of interesting books to read. The search I typed is ‘Stuff being made, how it was made, how history shaped its form’.
When I see lists like this I am surprised that they always live out school textbooks.
School textbooks are the kind of books that have the very precise purpose of giving you a foundation and roadmap to explore and understand a whole field of knowledge. And they're an industry that has been improving them on successive iterations to produce highly organized, comprehensive and accessible reading material through years.
For me the keyword in your comment is "also". What this word means to me is that, whatever faults textbooks might have, there is no alternative solution. They replicate a problem that all books have.
However, within the domain of textbooks, you still might find good solutions for these bias. These bias are mostly within history and social sciences and good and reputable Universities are renowned for publishing books in these domains that are quite balanced, comprehensive, deep and rigorous. Oxford and Cambridge are the 2 best examples that come to my mind. The U.S. has good Universities too, but after their government declared a war on science and knowledge I'd be carefull before trusting these universities.
This is pretty nice, I might use this if I'm looking to learn more on a topic.
I'm always cautious when reading nonfiction because it's hard for me to tell if the author knows what they're talking about when I'm not an expert myself. Using awards is a good metric. Maybe.
I am curious what the "score" for each book means. Is it calculated by giving each award a certain weight and adding them up?
Thank you! Wonderful idea, and very well executed. Useful discovery tool.
I understand your passion is non-fiction, but would be wonderful to have the equivalent for fiction, of which there is a great deal of excellent literature. Perhaps the same site could incorporate both fiction and non-fiction, with a filter?
maybe I am being too small brained black/white about this, but something feels amiss about the idea of vibecoding a solution to an AI generated problem. How can we get better, doing the same thing that made us sick?
I love the "books for dads who like Pavement" prompt! It would be great if you could incorporate the same experience of "browsing" random page of some of the results that are returned
I like very very much that there's no user voting system here.
I just recently read a really mediocre book with a 4.7/5 rating on Goodreads. Shows being sloppy is not exclusively a privilege for neither the AIs or producers.
I read at night in bed. I love how, with ebooks (first Overdrive, now Libby), when I hear about an interesting book, I can within 30 seconds search for it across multiple libraries, borrow it, and send it to my Kindle.
Serious question: can we define the top 5 parameters that distinguish human-generated non-fiction writing from AI "slop" writing (I don't like the term "slop" because it carries a built-in bias)?
Since AI output is based on the most likely next word/token given the context, let's consider how humans do this, i.e. choose the next word based on context.
Is one of the parameters the degree of "randomization"?
Are these other parameters?
- Specific experience?
- Recent exposure to relevant materials, i.e. paraphrasing?
- Borrowed knowledge from a specific resource?
- Writing style, i.e. choice of words or sentence structure?
I'm not sure how I feel about "Entangled Life" on the first page of books in "science". Isn't that slop? I mean, it's pop-science bullshit, and it's based on feelgood extrapolation from a thin evidence base. Maybe that doesn't quite fit the definition of slop, but it's not a signal of quality for the rest of the list.
As it happens, I had the same job as the author at uni. I shelved books at the biomedical and sciences library. As a life-long book lover, it was a dream uni job being surrounded by all sorts of historical treatises on medical arcana.
I've been really exercising my non-fiction reading muscles of late. I'm trying weaponise it as an antidote to AI slop and doom-scrolling. My last read was Designing Data-Intensive Applications by Martin Klepmann (the newly released second edition), it was a wonderful read!
The antithesis of AI slop is when you've created a unique idea from first principles and AI keeps misunderstanding your idea and spell checking it back to something that's widely popular such that you have to write a whole extra alignment subsystem to check that it hasn't ignored half of what you've wrote in your first principles document and just thought you were talking about something distantly related in its training set.
To work on something truly original in non-fiction in the future against AI pushing everything to the mean of its training set and fighting against that when doing research, as opposed to generating a whole book of slop from familiar ideas, is going to be quite the unrecognized effort of writing those books.
On a similar note: on Steam, the original Dark Souls game[1] is listed with the "souls-like" tag. That's not wrong, I guess, but just kind of a tautology. Grice's maxims[2] generally have such tautologies omitted in ordinary conversation as non-informative.
I participate book club for several years with friends, we casually navigate certain topics, and i can tell you human slop in literature is a real thing.
- almost every book try to stretch core idea into book size format
- unique ideas are rare people attack them under different angels
- a later phenomena : their believes almost predict entire book, outcomes etc, brainwash impact is real
so not sure, how ai slop is better vs book slop, at least with ai you can distill the idea, with the book, you have to spend 10-40 hours to digest average, absolutely non fresh ideas, that author brought in just to sell that book, otherwise it would be magazine article worth.
This is sadly correct/true. This is one of the many reasons why I militantly reject carbon chauvinism. Most of the potshots people take against AI can be easily retorted with “have you seen how bad the average human is at X?”
> how ai slop is better vs book slop, at least with ai you can distill the idea
Why don't you just use whatever LLM you have on hand to summarize every book you could read instead of reading it? The idea that ai slop is 'better' because you can distill the idea is an insane point to me because it treats books as a pure consumption-based concept, where the goal is strictly to finish a book and move on.
I think even the worst book is more valuable than whatever garbage an LLM writes because you can at least learn from the intent. Where and why the author failed, where they got stuck, the points they failed to make. LLM writing is a void, there's nothing to be learned.
In your experience, how do you determine book slop? For me I don't really stray from authors and topics I like, and non-fiction isn't really my go to outside of sci-fi, and mystery/law books. Like on the stretch point, I can tell that in other media but not so much books.
> And before you wonder, yes this is actually free. I am paying for the hosting and the API costs entirely because I just want people to find and read more good non-fiction books.
Just to note, you do have affiliate links so it isn't 100% out of the goodness of your heart ;) Not blaming you or anything, just saying...
It does not do the most important function well, giving me a list of award winning books that is easy to sort through. It is not clear what is a link and what is not.
It looks fine though somewhat generic. But it is not actually functional in a real sense.
I don't think we would have AI slop, if they were just trained on high quality media.
Less relatable AI, maybe, but not sloppy AI.
AI trained on other AI (i.e. distilling), can be better than the AI they were trained on, likely because of this reason. Each iteration can raise the bar.
As apposed to the deep spiral of generating shiny attention crystals for humans, and then training on that.
The slop isn't a property of the LLM but the workflow it's utilized in. Just now I asked ChatGPT about the linguistic effect where a CRT could be described as a giant flatscreen, a landline could be described as a wired cellphone, or the original Rogue could be described as a roguelike. Unfortunately it doesn't know what that's called either, but that is a non-slop use of an LLM, like a semantic search engine. On the other hand if you print LLM output into a book it's almost certainly slop no matter how the model was trained.
Obviously, you can handicap any model with an indiscriminate dataset. Are you training on slop? Why?
More data is not automatically better data.
Just three (of many) advantages of quality data selection: (1) It gives a clearer signal. (2) It, ironically, reduces the complexity of what needs to be learned, since slop adds its own complexity. (3) It reduces dataset size which means more learning per watt (and per just about any other cost). The advantages compound.
Humans are no different. Children surprise with their ability to absorb sophisticated relationships and skills, while people who had bad examples struggle to recover.
99% of human effective intelligence is higher quality cultural knowledge learned as children. None of us had to spend centuries deciding whether zero, negative numbers, the square root of -1 are numbers. Slop + quality is not quality.
It's sad now to see a university library where students sit among endless shelves of amazing books, sitting on laptops, almost all of them with ChatGPT open. The library has become just a place to sit and open an LLM.
It's especially sad because on of the things those LLMs are best, almost purpose built for is to tell those students which books to go open, where to find nuggets people haven't bumped into for years, to make cross-connections that would take a PhD a decade to find.
The most sad part of all is to think how little of that knowledge is digitized and available to be referenced by ChatGPT. Digitizing old books is one of the highest things you can do as a human for “good for the world”.
Nice site, but most of the top biographies are US figures: Edgar Hoover, Frederick Douglass, Lyndon Johnson etc.
I'd like to see more non-US biographies.
"Omit internet tropes." (the stopped-reading-at bit)
"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."
Edit: it looks like your account has been posting quite a few flamebait and/or unsubstantive comments generally. Could you please not do that? It's not what this site is for, and destroys what it is for.
I think my issue with this project--and so many other similar ones--is that the provenance of the code does undermine the intention. If a project purports to be about quality, then knowing that the creator abdicated some of the responsibility for creating the thing they ostensibly care about makes it harder to put faith in them as having high standards elsewhere.
Perhaps I am just old-fashioned, and vibe-coding is something that can be done with full focus and care for high quality, but I remain unconvinced. This is a project that needs to be cared about sincerely to be trustable/meaningful/useful, and the approach taken casts doubt on that.
Not to be intentionally contrarian, but I have thought a lot about this:
I think good fiction is actually the antithesis of AI, fundamentally. LLMs are really only capable of combining different things into output. This can give you some creative results, but it is ultimately kind of limited. I cannot imagine that prompting an AI a thousand times would give you something truly original in the way many high-quality fictional stories are. The magic and absurdity of life is a necessary thing to create such works.
> LLMs are really only capable of combining different things into output
> The magic and absurdity of life is a necessary thing to create such works.
These two points seem slightly conflicting. Why if a human is truly more creative than AI, must it experience "the magic and absurdity of life" if it can be creative without "combining things into output"? Is a creative human then just not splicing and recombining their experiences into their creative output? And therefore, if an AI had more data that mirrors that experience in some way would it be sufficiently creative to create something "truly original"?
My view is that intelligence and creativity is all just purely recombination. AI can do that well just as humans can. The gap is that the data humans have access to is much more analogue, emotional, and grounded in the real world compared to the data that AI is currently trained on.
> And therefore, if an AI had more data that mirrors that experience in some way would it be sufficiently creative to create something "truly original"?
An important distinction seems to lie in a human having original experiences to draw from, while an AI has only "data that mirrors" experiences.
> while an AI has only "data that mirrors" experiences.
True for a lot of current AI. With more robotic bodies, loads more access to cameras/audio/satellite/science measurements, an AI can have original experience on a level that can only be described as god-like compared to humans.
You could argue AI already partly has access to these measurements. You could easily tap into smart sensor data etc, same voor video/audio etc.
The part that AI can't (yet) truly replicate is 'feelings/emotions'. Experiences such feeling dread, loneliness, happiness. Or being anxious, nervous or having goosebumps. You can describe all these states but they are differently experienced for each individual and also dependent on each unique situation.
Ofcourse this also then comes down to how you exactly define emotions.
We don't know that Claude doesn't feel dread when it looks at my code. To be fair, these emotional circuits are the result of evolving in various environments. AI might well have completely new emotions that we just can't comprehend.
And at this point you’re far off into science fiction, and beyond reasonable discussion of this.
Even then, it’s not really clear to me that an android with the exact appearance of a person will have the same experiences. They aren’t biological creatures with millions of years of evolution behind them.
If it has a subjective viewpoint, i.e. sentience. Otherwise it still doesn’t actually experience any of those things, and can’t tell actually you what it’s like to experience those things. It can give you the data readings though.
Wouldn't it be funny if they missing thing is actually the body, and having one to give the subjective viewpoint something to center on directly leads to emerging sentience?
Though I think there's at least one more important thing missing, namely the ability to form memories - to incorporate recent context into the model's weights. I don't think it's a coincidence that this is a very complex process in humans as well (involving sleep and dreams as neccessary mechanisms).
ChatGPT has had more conversations than you or I have. And it's conversed with people from a wider range of backgrounds in a wider range of situations.
Billions of conversations. Zero lived experience.
The difference is huge. An LLM conversation can contain within it a batshit crazy model of the world and still produce a plausible stream of tokens. No skin in the game. If you don't test your ideas against reality, you don't have a model of reality; you have a model of written human thought. It's critical to understand the difference.
And? Can you tell me what it feels like to eat a hamburger by asking a thousand people about it?
I think this is actually a tough one. Yes, I agree that an LLM can never experience this. But -- when a real human needs to translate that experience back into words, it can actually do so just fine.
In other words, when a real human needs to translate an experience into words to share it, A LOT of information is lost. The LLM might actually be better-suited for the the task in your example, even if the point you were driving at stands.
I don't think it's all that difficult. Language is not a 1-1 match to reality. Ergo any entity which is inherently language-only, will fail to accurately describe the experience.
Not if the only result you can judge is also language-only, and the entity has access to a large corpus of descriptions produced by people who have had the experience.
Then you're just comparing descriptions, and you're also dependent on the language skills of the experiencers, and you're still stuck with the problem that language is not a complete map of reality. So an LLM may be able to give you the optimal description of a feeling, on par with what a person would give you; but it won't be equivalent to the actual feeling.
>and you're still stuck with the problem that language is not a complete map of reality.
Definitely agreed, but at least when talking about experience we've been stuck in this position for most of human history. How do I know what it's like to eat a hamburger? Perhaps, like the cute visual in Ratatouille there is a deep inner experience but (also like in Ratatouille) when I need to communicate that experience with another person I'm stuck relying on language, and the language is found lacking.
I know it's just a kid's cartoon, but interestingly Ratatouille was making this exact same point -- he really lacks any meaningful way to convey his inner experience. It's locked behind language. The film tackles this by showing an Remy's complex and interesting visual, but his friend only gets the fuzziest, dullest visual when he attempts to communicate it.
Remy's immediate rat & family peers actually _never_ get to understand his experience, and instead later help him out of love an allegiance.
https://www.youtube.com/watch?v=xizttM_Cbuc
Does it matter (why and how) if you can't tell the difference, no matter how hard you tailor the question towards that end?
It's basically the Turing test.
That's what we all did in school, didn't we?
Nobody did any original research, you didn't discover gravity, and you didn't participate in any historically significant events.
Yet we say that people who have been through education know about such things.
That is descriptive knowledge. Not firsthand qualia.
You may know every detail of the American Civil War, from every book ever written on the topic. You may read every journal from every person that wrote one about it.
You do not have firsthand knowledge of what it was like to participate in the war. Any assumption of that knowledge is a (false) generalization of what you think a person “would have thought or felt” based on contemporary ideas of the personality, psychology, etc. Reading a journal entry about a soldier seeing his comrade killed is not equivalent to actually having that experience yourself.
So only people who have fought in wars can write novels about fighting in wars?
If the aim is to have actual knowledge of the experience, then yes.
Of course not all works need to be this way, and many, many books are written about war by people that haven't been in it.
The broader point is that LLMs cannot have original experiences.
But that's purely a matter of having sensors?
Stick a camera and a touch sensor on the machine, now it is experiencing the world.
Besides that, there is a lot of value in redigested experience. Quite a lot of history is only really examined in a larger context, where the person describing it cannot have experienced all the original events.
Is a human life, as experienced firsthand, the same as a first person camera recording it?
No, it's the same as a camera recording it and digesting it.
Ah, I see the goal posts shifted. I thought the topic was whether an AI could "give you something truly original in the way many high-quality fictional stories are." Which was then followed by the idea that you cannot have "truly original" outputs unless you also have the corresponding original experiences.
Implication is that virtually none of the novels ever written which include wars are actually original works, since vanishingly few novelists have been in active wars.
Maybe originality just doesn't mean much then?
He probably could tell you something about it, based on other people's experiences. Doesn't mean he would actually know what he was talking about.
He could tell you a collection of descriptions. Not actual knowledge firsthand.
Does reading an infinite number of cookbooks mean you know what the ingredients in each taste like?
Yes, of course.
That’s a pretty lacking theory of knowledge. How can you know what taste is, if you’re never tasted anything?
If you read a thousand books about living in New York City, is that the same thing as the experiential knowledge that comes from living there in person? I certainly don’t think so.
keiferski didn't ask for a theory of knowledge. You're moving the goalposts.
That's literally what the conversation is about: what is knowledge?
You cannot claim that descriptive knowledge of something is equivalent to experiential knowledge of it, unless you have a theory of knowledge explaining why this is the case.
You asked for a description: "Can you tell me what it feels like to eat a hamburger"
I am sure I Fable (or GPT-5.6-Sol) could answer this question very well. It can probably even tell us what it feels like to eat a hamburger for the first time when you spent the first 18 years of your life as a vegetarian.
I don't think you understand the premise of the question.
Asking another person or Fable what it feels like to eat a hamburger will give you a description of what it feels like. It will not give you the actual feeling. You are merely repeating words that someone else has told you. The description of something is not equivalent to the experience that comes from doing it yourself, as an experiencing subject. Reading a war journal about being in battle is not equivalent to the actual experience of being in battle.
To illustrate the point again: if we gathered ten people that all spoke different languages and gave them a hamburger, they would all gain the knowledge of what it feels like to eat a hamburger, even if none of them are able to communicate this to the others in language.
I understand the premise. My answer was meant to illustrate that, no, the 'original experience' is not needed, and that AI can produce a good answer relying only on '"data that mirrors" experiences'.
"Is a creative human then just not splicing and recombining their experiences into their creative output?"
I do think so. The difference is that, even apart from synthesis, the creative human knows how to filter out the shit ideas from the clever, subtle ones—LLMs seem to just run with the first…um, thought that… um, comes out of their head.
I suspect knowing that you've on to something special probably does require "human experience", woe, melancholy, "the human condition", reflecting on the brevity of life…
> My view is that intelligence and creativity is all just purely recombination. AI can do that well just as humans can. The gap is that the data humans have access to is much more analogue, emotional, and grounded in the real world compared to the data that AI is currently trained on.
Yes!
Aside from the more fundamental questions about the nature of cognition and what constitutes intelligence, there's a practical matter: humans have life to learn from. We have a lifetime of persistent memories, an identity, a social life, physical interactions with the world of many qualitatively different kinds, and we're taken on a lifelong journey into different studies, workplaces, and living environments by the biological need to survive.
Even if human creativity is in some sense also recombinative at its core, we have a very different set of inputs from LLMs. Even under the (almost certainly false) assumption that human condition is basically similar in structure to an LLM, we should expect some differences in outputs based on these profound differences in inputs.
>These two points seem slightly conflicting. Why if a human is truly more creative than AI, must it experience "the magic and absurdity of life" if it can be creative without "combining things into output"?
I agree, yet I also think AI is extremely limited in terms of new creativity.
My guess is that humans operate on some deeper level. Much like the grokking model example where it easily trained on the training data, but only later did it really start to understand the deeper problem space and 'grok' the problem, I think LLMs are doing something similar with human thought. LLMs can throw more compute, but they are still lacking some introspection/drive that leads to the better creativity of humans. Grokking this from the training data might not be possible, or it might take some fundamentally different approach, or even an imperfection in humans that LLMs don't have (ego/drive?).
This isn't to say that LLMs could never do it, but the cost to do so might require magnitudes more scale in training times and size. Or it could be in the next model released. Or maybe they already can, but that gets trained out in favor of overall more general correctness accross all domains.
I think you're just neglecting the fact that LLMs only exist in the virtual world, whereas we're interacting with a world that's richer and has much more to offer. The virtual world is still just a mere subset of our world, but whether we want to perceive it as such is up to each individual. It certainly feels like you're conflating digital with analog, and a digital signal can still only approximate an analog one.
I still doubt that we have the knowledge to describe all of our human desires as mere results of our brain interacting with the nervous system either. There are many processes going on besides analysis and synthesis. If I plan on doing something extraordinary, I don't only sit there, think, and try to rearrange some pieces from memory. I'll go look around, touch, taste, re-explore and expand the perception my environment has to offer. Further, my perception can vary by the minute and I could even try to bring in a touch of chance whenever I feel like.
Wittgenstein's "whereof one cannot speak, thereof one must be silent" might be fitting here, yet language still continues to evolve.
"…LLMs only exist in the virtual world, whereas we're interacting with a world that's richer and has much more to offer."
I have a gamer friend that believes, having played 100's of 1st-person games, that he has had rich experiences because of it.
I reflect back on a day my family and I set out to hike through Arches National Park (still a Monument at the time, I believe) and the heat was such that my oldest daughter seemed in danger of going into runaway heat exhaustion. I began to panic as we were so far along on the hike already that even returning to where we parked seemed to pose danger for her.
There was little to no "coverage" (shade) along the trail and the heat and sun were intense that day—we had passed almost nothing. Finding the smallest of bushes to provide minimal shade for her, I scouted ahead on the trail to look for some kind of larger area where there would be enough shade, perhaps a breeze even, to allow her to bring her core body temperature back down.
Fortunately I found just that. (And it turned out that, in the news later, two hikers within a few hundred miles of us had in fact died that day as it was a record heat wave for the area.)
I think somehow that me and my daughters experiences that day, when actual death was perhaps on the table, was something a "virtual environment" can never really offer.
Is a creative human then just not splicing and recombining their experiences into their creative output?
No, they aren't. I'm not sure why it's such a popular narrative that all art is just some kind of recombination of existing things; it's not. Actual lived experiences are unique and impact what is created. Novels written by people in post-Revolutionary France are not the same as novels written by post-WW1 France. Events in history shape culture and introduce new ideas and experiences that weren't there before.
The gap is that the data humans have access to is much more analogue, emotional, and grounded in the real world compared to the data that AI is currently trained on.
That's the problem with your argument, right here. Emotions are not merely portable "data," and framing this way is a problem from the start.
I don't think you get human experiences without actually being human. The most perfect LLM copy of a mind is just the copy of externalized output.
It's often said that creativity involves reuse because that's clearly part of the process. It's like a warning against another narrative, where a reclusive genius outputs great works made from scratch, without being part of culture. Yet there's some truth to both. The thing is to be inspired by the culture and then be defiantly original and work away stubbornly in private and also talk to your peers and your audience. It reminds me of how we used to play as kids: oh you're doing that thing? I'm gonna do that thing too ... when you see the big reveal, you're gonna be very impressed! I'm doing it differently! Look at how I'm doing your thing that you came up with!
"I'm not sure why it's such a popular narrative that all art is just some kind of recombination of existing things; it's not."
I would like to see art that was "created in a vacuum" so to speak. I am not aware of any. A first time poet who had never read any poetry or prose or heard a tale told? A painter who had been blind since birth and just handed a canvas, oils?
(To be pedantic, the person you quoted said art was recombining their experiences which you took as existing things.)
"…truly original in the way many high-quality fictional stories are."
The older I get, the more "nothing new under the sun" returns again and again.
"I will say nothing new!"
- St John of Damascus
[dead]
Agreed! I find AI to be pretty decent at documenting and explaining technical stuff, which is a subgenre of non-fiction.
I once tried telling it to write a fantasy novel for me because I didn't know what to read next. Gave it some examples of books I have enjoyed in the past so it would have an idea of my taste...God was that awful. Absolutely terrible idea.
AI is not coming for fantasy authors jobs any time soon!
Models are increasingly showing their ability to extrapolate into the human unexplored (math proofs being the most apparent). What gives you confidence the absurdity of life is uniquely difficult for models to source?
Mathematical proofs are hard to come to, but easy (or “cheap”) to verify. Also having a math proof processed through a transformer to derive an output that matches existing proof patterns kid of makes sense. The former allows you to generate hundreds or thousands of iteration and cheaply verifying its correctness until you hit the right randomness level. Generating thousands of fiction books then verify them doesn’t really work.
Most good fiction authors are trying to communicate a story that’s forming in their minds. Even when they get stuck, they are usually not stuck because they just don’t know anyone to ask. It would’t be their voice anymore. Some do seek input and those might experiment with llms for ideas. Using llms for research, technical questions, grammar, synonym sentences, word selection, etc makes sense too.
I’d agree with OP for people generating full stories through llms. While I say that, I know a very young person who told a while back (like early ChatGPT 4 time) they use it to write fan fiction for books or characters they like.
Being able to write a series of logically correct sentences down is not the same thing as intuiting human emotional experience. Or even related.
Math seems entirely unrelated to the human experience to me, at least in the sense of “this thing happened to me and I will transmute it into a work of art.” Math is pure abstraction.
I will readily admit that I didn't read Dostoevsky, and I don't continue to read Dostoevsky, because I've compared him against a swath of other fiction writers and authoritatively determined with some objective reproducible standard why his writing has been impactful and changed my life. I heard about Dostoevsky from a lot of people talking about depression when I was depressed, he resonated with me. Then when I came back to him half a decade later, having resolved more of my issues, I was able to evoke similar feelings but through a more mature lens.
Do I reflect on the objectively correct structure of the story? Do I "agree" with everything Dostoevsky is saying? Not really.
I don't know if this offends literature people or not. But this sure as hell is incompatible with the AI bro paradigm in which all thought can be objectively reduced to some shared standard to benchmark on.
My position will be read into as deep irrationality from many opposing positions, but it's fine.
Now, none of this doesn't "preclude* the ability for AI to have the same kinds of network effects. Of course it can - that's what I've been saying for the longest time. But don't confuse, becoming integrated with society, with this ideal of superhuman X.
Now of course, you could argue, "there is no such thing as anything else besides becoming integrated with society. There is no other standard that works." Very good! Darwinian-adjacnet positions are interesting. This is a position - but then don't claim it's equivalent, or god forbid, the exact same, as the "super AGI" ontology. These require fundamentally different assumptions about what justified what, and an attempt to try and pretend otherwise is sophistry.
I have also thought about this a lot but have less developed views on fiction (short version: I don't currently want to read an LLM-written novel, but I expect that will change in the future and there will be some I enjoy)
But I want to say that non-fiction books are a lot more resistant to LLMs than you might think. I'm reading two non-fiction books currently. And both (https://www.goodreads.com/en/book/show/17801.Underground, https://www.goodreads.com/en/book/show/44824581-a-k-pop-live) could not have been written by an LLM. One is based on in-person interviews that were hard to set up, and the is based on a log of sources a number of which are going to live events.
Not every book fits this pattern of course. But even a more straight forward book (https://www.goodreads.com/series/139929-designers-dragons) will often rely on sources that are not easily accessed by an LLM, they may not be digitized, or they may not be accessible in your common digital collections.
I 100% think LLMs can do good synthesis work, don't really think we have the tooling/checking in place for them to write a competent non-fiction book yet, but that will certainly happen, but I still think the best and most interesting non-fiction remains firmly in the realm of humans.
(Also, a lot of non-fiction is memoirs, travelogues, etc. which LLM can't really replace unless you're dictating to it.)
While you are right that humans can probably find some original ideas that are not just combinations of inputs, the universe of things that humans will find interesting to read is not that vast, and that’s why there are only really like 7 types of stories, and those are the ones that survive through a process of natural selection.
Everything else is just not that interesting to a human.
it's a real question. I've seen a lot of people stating that the human mind "is more than this" as if it was obvious
I would say the inherently chaotic and continuous nature of the human brain can create novelty in a fundamentally different way than LLMs can.
Like a human can reach into the chaos of their own emotions and pull something unique out. There is some layer of the human experience where ideas can ferment and change based on all your memories is weird ways, and LLMs completely lack this, which makes things LLMs create very blunt and obvious. Like there is no nuance to anything they do.
Well since human culture has developed from very limited stone age culture into what we have today, we must conclude that at least _some_ people are capable of generating something that is genuinely new.
So now we'll have to investigate what "original" means?
From what I can see, AI is absolutely able to create new work. Things that are not in the dataset, but look like they could be. This is pretty much what generalising means, they can make a wide variety of things that aren't actually in the original set but can hide in that set pretty well.
When you are an original piece of culture, what do you mean by it?
Can an AI write a story about living as an artist in NYC in 2026, the cultural environment, etc. without that information being added to the dataset immediately as it happens?
It seems pretty obvious to me that there are cultural trends which are happening instantaneously and aren’t merely some sort of consequence of past historical data.
Ideally the AI would get none of its world knowledge from its model weights and would rely entirely on context. This is just an ideal though, we haven’t figured out how to completely separate world knowledge from the model.
But how could something be a cultural trend without some sort of anchoring in history?
I don't see why a computer (or a person) couldn't make up a plausible sounding story about a NYC artist of the present day? You would just reference things that don't change that much (artists having a boheme type existence, wild inequality in who can live off their art, NYC being a sort of place where people go to become successful).
you have a bad imagination then
> LLMs are really only capable of combining different things into output
In terms of how writers think about creativity this doesn’t reflect reality- there’s the old writer’s saying that there are only 7 stories (and other variations of this idea) and that writing is about creatively remixing these well-worn ideas.
The LLM’s capability to write acceptable fiction and nonfiction is coming soon if it’s not already here. (I think right now it still needs some high level input from humans, depending on length and topic)
It’s clear to me that, especially as LLMs get better and better, there’s only one real difference- it’s that I don’t want to hear what an LLM has to say. I’m only interested in what other humans have to say. I feel the same way about AI music- even if it sounds ok, even as good as the kind of unoriginal pop music that’s not AI made (but is a kind of it’s own slop) or a bad committee made hollywood movie, it’s still fundamentally more interesting than something that’s been generated. Even if it’s 100% fiction it’s still based on a real human’s life experiences.
This is a bit like saying “there are only five fundamental flavors” and everything is some combination of them.
Maybe, but that has basically zero effect on the experience of actually eating food. Just because the form may be the same doesn’t mean an LLM will be able to supply original content.
I’m not sure I understand the analogy. You could say that the reason why artificial flavors work in food is because we know how to recreate flavors from first principle chemistry. The dislike of the artificialness of cheetos also doesn’t stop many many people from eating them.
I would say human made slop writing isn’t any better or worse quality than LLM writing.
Knowing if it was made by a real person makes a difference to me, but I also know it won’t to everyone.
Just because the form of a story is the same as another story does not imply that the content is interchangeable.
I could write you a story right now about being an expat living in Poland in 2026. It’s unclear to me how an LLM would have the cultural knowledge or experience to write such a story that is actually accurate and up to date, without already having something like my story already in its dataset.
Yeah but you would just be doing the same thing: mashing up some known information about humans and Poland. You would pick plausible names like Marek or Piotr, plausible places like Warsaw or Gdansk, and some random flavour of story that is either completely general across cultures (hero's journey) or some local tale from Poland.
Shakespeare did the same, so I wouldn't blame you for it.
I am an expat in Poland, that’s why I used it as an example. The point being that I have knowledge of that experience that is not written down in an LLM dataset.
Which would you rather watch:
Superhero Wars Trek 42 - the latest franchise installment, billion dollar Hollywood budget, financed and produced by billionaires, audience tested committee-made script, lots of people involved, designed primarily to push merchandise.
Personal Story - a movie about someone's personal reality or fantasy, something typically not made by studios because the story isn't profitable to a wide audience, painstakingly generated and edited by one person using AI, designed primarily to tell a story.
I put up a different strawman because your strawman is too beatable. As humans, of course we're all interested in what other humans have to say. But you haven't explained why you feel people can't use AI to say things, or why AI output can't be based on a real human's life experiences.
Personally, it seems like movie studios are the ones with only 7 stories to tell, and I've seen the movie and its sequel, the prequel, the requel, the remake, the gritty reboot, the musical, and the animated series. I'm excited that AI will let people skip the studio gatekeepers to tell some new stories for a change.
[dead]
Exactly. Shows us what is true in the sense of the experience of being a human. It is resilient against slop because good literature has experiential depth. Maybe LLMs learn to emulate this some day, but I doubt it. Not enough money in getting it right.
that is how humans do it too! they combine things, ideas.
you can make LLMs do that to
Thank you for sharing this.
I do think book prizes are a better-than-average signal but I previously volunteered with a book award. I will caution that pretty much every publisher mass-submits these books for consideration in every remotely relevant prize. It is a cost of doing business (similar to how photographers pay to enter photography competitions to try and win the "award winning photographer" title, or businesses submit dossiers with consideration fees on why they're one of Michigan's top 100 places to work).
There are often so many books and so few willing qualified readers that which books get an award can either be completely arbitrary, or comically easy.
For example, the NCR Book Award in your book corpus faced a big scandal when it was revealed that the judges did not read the books themselves. [1] The PROSE award is so comically large that an ordinary category finalist or win is usually overstated in prestige and value.
[1] https://www.theguardian.com/news/2013/may/19/literary-prize-...
Awards are also a product of their time, representing biases among the judges, literary preferences from the readership, and straight up nepotism/favoritisim/elitism. Look at the Hugo/Nebula/Locus and you'll find a lot of so-so winning work throughout the years; especially when you look at the other nominees.
Same thing with beer and wine awards, some are good but most are just a paid promotion lottery.
Which are the good ones in your view?
People’s tastes in beer and wine vary so dramatically (and there are so many styles) that it’s not clear to me that you can actually have a “best of.”
The beer and wine I enjoy the best my wife finds completely unpalatable. For wine, I enjoy a dry Grenache; Mas de Gourgonnier is a good one. For beer, I enjoy a beer with body, a malty profile, and balanced hops; there are many mass-produced styles that fit this, but I like Sam Smith Organic Lager, and Leffe Brune. I’ve also brewed many beers myself and one of my all-time favorites was an all-grain formulation of a honey ale [1].
Such is the nature of beer and wine (and HN) that someone is going to “disagree” with me. That’s fine. Just go try some things and see what you like.
[1] https://en.wikipedia.org/wiki/White_House_Honey_Ale
I would bet that how your brain stores information that you read from long-form text is very different from how it stores information you acquire from chatting with an LLM. When I read something challenging or new to me I spend a lot of time thinking about how what I'm reading matches my own experiences or knowledge. Although I'm a fairly fast reader, it often takes me a long time to get through difficult pages since I have to stop and think about what I'm reading. I seem to be doing a lot of integrating and reorganizing my thoughts. When interacting with LLMs it feels a lot more like I'm just receiving knowledge passively and I don't think it gets integrated as well. Not sure why this is and its somewhat counterintuitive since I don't think I'd have the same experience with a human tutor.
> When interacting with LLMs it feels a lot more like I'm just receiving knowledge passively and I don't think it gets integrated as well.
This sounds like it's more down to how the individual uses the tool. I am not someone who has even been particularly good at reading > learning the thing. I've only ever been capable of learning by doing and with capability to interrogate on the points I don't get. LLMs allow that in a way that is just not feasible with any human being whose tolerance of me would diminish rapidly.
In most cases, I will be writing down my understanding as I would in isolation from a primary source, building flash cards and actively practicing what I have learned, only with more capability to interrogate on the points I have difficulty understanding or need clarity on. Effectively, I am doing the following in a capacity I personally never had via any other means:
> I seem to be doing a lot of integrating and reorganizing my thoughts.
If you are using it as a slot machine of knowledge, and going from receiving > doing with no intermediate step I see how outcomes could differ.
Seconded. At a minimum, you can literally tell the agent to guide you but not point you and only help when you are stuck if you're actively trying to learn something. It's been a help for me to bridge a couple of gaps I'd struggled to cross before (mainly hardware things).
> This sounds like it's more down to how the individual uses the tool.
I genuinely agree with a lot of your comment but lines like this drive me crazy. Anytime somebody has a critique of AI, so many people hand wave away with “well you’re doing it wrong.” If we can just chalk up every negative thing with AI to “incorrect tool usage” then we’ll never honestly assess it.
I didn’t say they were using AI “wrong.” That’s a stronger claim than the one I made.
Their assessment was:
> When interacting with LLMs it feels a lot more like I'm just receiving knowledge passively and I don't think it gets integrated as well.
My point is that the passivity is not necessarily an inherent property of the tool. It depends heavily on an individual chooses to engage with the output.
Saying LLMs make learning passive is a bit like someone else saying books make learning passive because you can skim them without thinking. When we all know instead of reading a book passively you could in addition annotate it, question its claims, summarise it, test yourself, and ultimately apply what you learned. The medium permits both. The same is true of the output of an LLM. You'll get out what you put in and nothing more.
You can use it like a slot machine, repeatedly pulling the lever for another answer without reflecting on any of them. But the fact that a tool permits shallow engagement does not mean shallow engagement is the only option available to you.
That also doesn’t mean every criticism of AI should be dismissed as “incorrect tool usage.” AI has real limitations and deserves serious criticism, and I assure you, I am not popular with the AI hype bros for this reason. However we still need to distinguish between limitations of the tool and consequences of how someone uses it. Not every potentially negative experience is rooted entirely in the technology itself. Sometimes user behaviour is a significant part of the cause.
I listen to audiobooks while I walk and hike, and sometimes when I recall some particular thing I've learned, I can also recall where I was hiking when I heard it.
I think there's a lot to learn on how we really process information.
I once heard a lecture about „learning how to learn“. There are already a lot of well-researched things one can exploit about the way we process information. What you describe is known as retrieval cue. Our minds seem to be associative^1. It also seems a lot of the learning/processing happens after the actual learning process, which is why resting is important. Spaced repitition is a further method well-proven in practice.
^1 In the early days of AI, the concept of Hebbian learning was popular („neurons that fire together, wire together“). Its implementation is even simpler than gradient descent used today, but never caught on https://en.wikipedia.org/wiki/Hebbian_theory
Perhaps not the same one, but there's a fantastic course by Barbara Oakley by this name. I viewed it quite a while back, but the modern incarnation seems to be [0]. In a similar vein, the book "How we learn" by Stanislas Dehane is also very good!
[0] https://www.coursera.org/learn/learning-how-to-learn
Same here. I always marvel at this. These recollections can even be years later, and the memory of the place in which I heard the remembered thing is vivid.
There's a podcast I listened to as it came out when I was a teenager. I'm now relistening to it over a decade later, and I distinctly remember walking home from the bus stop with it coming through wired earbuds from my ipod.
That's spatial memory! I have long walk-and-talks with friends, and sometimes I forget minor details of previous discussions, but not where we were when it was mentioned.
A lot of my friends can't do same trick. But the fascinating part is that it still works for them! Usually they can recall whatever I forgotten after I described the place we were in.
"when events are represented in memory, contextual information is stored along with memory targets; the context can therefore cue memories containing that contextual information" https://en.wikipedia.org/wiki/Context-dependent_memory
i read a school-assigned book during the loading portions of a certain computer game. the two stories are completely interlinked in my head.
What you're describing is very close to what education theory calls "assimilation" and "accommodation". When we assimilate knowledge, it just fits into our schema of understanding. When we accommodate knowledge, we need to change our schema, and that is where the feeling of being challenged (and often the feeling of profundity) comes from.
For example, a child learns that "foreigner" means someone from outside their country. Then, when they're 11, they go on their first holiday abroad and realise "Wait! _I_ am a foreigner here!"
So, maybe one way to frame what you're saying, is that LLM output tends towards being easily assimilable.
> When interacting with LLMs it feels a lot more like I'm just receiving knowledge passively
It's... almost an oxymoron, and very different from my experience.
In college I used mathematica a lot while taking linear algebra. I ended up having to relearn a lot of that math later since I never really understood it at a deep level, although I was able to use it and apply it under class conditions. I feel like a lot of my LLM derived knowledge is similarly superficial.
That's absolutely right. Math is difficult to understand without writing it down on paper. As one of my professors said, "something magical happens when you write down a problem from scratch and work it out". LLMs make people believe they are improving their productivity and that may be so, but it also removes that magic and removes that deep learning.
Any such bet would be grossly misguided given how different people acquire information differently. Just because you don't like using LLMs doesn't automatically extrapolate to a general realisation about the human brain or how one approaches knowledge.
Your subjective experience matters in this case. Almost completely.
I like using LLMs and one of the main things I'm working on would be impossible for me to tackle without them, but I do think they cause some skills to atrophy despite being a big productivity booster overall.
Thank you for this. I sorted the "technology" and "science" sections and saw a few excellent books that I've read and a few that I would really like to read. This motivates me to start setting aside "reading time" again every day, since I've lost that habit.
The recent "society & culture" books gave me some good book club ideas.
Bug report: filtering by "award" appears to be broken for some awards. If I select Pulitzer or National Book Award, no books show up, but I can find books with these awards by browsing.
Awesome, that's what I was going for. I originally made this for my own use and got kind of obsessed with building it out once I started finding unfamiliar books I enjoyed using the semantic search and category browsing. Glad to see it out in the world hopefully doing the same thing!
One thing that would make this information even more useful and searchable would be a JSON dump of your dataset. You could serve it statically on S3 or similar, to keep hosting costs to a minimum.
Of course, having created a cool thing generates no obligation for further work on your part! I'm grateful that you did this work and made it available for free.
Did you build your semantic search index based on the full text of the books, or just the reviews and descriptions?
Already tried to do this in a basic form. Advice from experts would be welcome as it's my first time trying to make data and an API available: https://book-prize-index.vercel.app/data
Also I just had to upgrade my Vercel plan due to traffic, which is very welcome and appreciated, but if anyone wants to donate, that would be really helpful! There's a Stripe link at the site: https://book-prize-index.vercel.app
Donated happily. Long-time book lover and am always looking for my next favorite read.
Thank you for building this!
Oh awesome, you're way ahead of me!
The JSON downloads quickly and had all the info I needed.
> statically on S3 or similar, to keep hosting costs to a minimum.
Oxymoron. S3 and its clones are one of the most expensive ways to host files.
I recently did this! I fixed my phone habit and my TV habit, and I just read no, the family has silent reading parties instead of movie night.
Here’s what I did: 1. Install Jomo on the phone. There are other apps, but basically it makes me wait 5s before I can open a brainrot app, and then I can only use it for 5min, and every time I do this I have to wait an additional 5s. It resets at midnight. This introduces the necessary friction.
2. Paper books. The house is now full of paper books. Book seems interesting? Buy it. Books everywhere. Real books.
3. The actual habit. I stack habits. Right now I have a daily workout routine. First thing I leave the house and either run or go to the gym. I bring my book, and afterwards go to the cafe across the street for a coffee. At night, I read before going to bed. Once you finish a few books this way, you’re in. TV now is actually hard to pay attention to, and I feel kind of… dirty doomscrolling now (like eating junk food, it tastes good but you feel gross during the experience).
If you want something to start with: https://bookshop.org/p/books/true-grit-charles-portis/5ac454...
I bought a tiny e-reader [0] and carry that as well as my phone. When the temptation to doomscroll triggers, I pull out the e-reader instead and read a book.
I tried attaching the book to the phone as their marketing suggests, but that didn't click somehow. Having both out means I would check the phone and then not look at the book.
Working so far, but it's only been a couple of weeks, so let's see.
[0] after reading this HN post: https://news.ycombinator.com/item?id=48662381
Yea, for me my Kindle just didnt work. I always have had one since they were first released, but it was always hard to keep the habit between book.
Having a nice library of paper books, for whatever reason, has worked a lot better for me. I think it’s passively seeing them around the house and in the background always considering what is next.
This last round I couldn’t decide what to dig in to after I’d finished True Grit, so I spent a few minutes browsing what we had on hand and it was just… a very pleasant experience.
Also, unless you pirate ebooks or have access to a library that offers them, used paperbacks are actually much cheaper than buying ebooks.
Yeah I pirate my ebooks unless they can be bought DRM-free (which a few can). On principle, because I want to actually own the things I buy.
I used to have paper books. But I found they were just a massive pain in the arse to carry around, and move. They became very expensive, very inconvenient wallpaper. I gave them all away to a charity shop and bought an e-reader instead. Now I carry around 250 books in my pocket wherever I go. It's liberating.
Thanks for sharing jomo. My hack has been to read on my kindle while standing or even walking short repetitive steps (in surroundings or patterns where I can't hurt myself). This has been a game changer while waiting for kids soccer games to get done, lines in the mall, baggage claim etc. It also helps me avoid sedentary reading while sitting.
Suggested additions: the Axiom Business Book Awards, Library Journal Best Books of the Year and Booklist Magazine's Editor's Choice Awards. There's also a few popular substacks that have big sideshows in regular book reviews, but I wouldn't even recommend my favourite for a public aggregator like this.
Original author/creator of the site here - thank you! Will add the Axiom one. I am on the fence about whether/how to add "book of the year" type lists as they are somewhat distinct from book awards, but I do think that would probably be the next step to get more books in the corpus.
And any other ideas that HN readers have for awards to add would be welcome. Currently it's probably too history-slanted since I'm a historian and knew those awards better.
It may be hard to parse the lists but if they're still available somewhere online under the current administration the past annual State Department, CIA bookshelf recommendation, Army Chief of Staff, Navy CNO and Marine Commandant reading lists may have some valuable additions. To my knowledge there aren't any analogues to them in the rest of the Anglosphere but some LLMs will probably point out ones I haven't thought of.
I can point out that the UN agencies' respective reading lists for their staffs' professional development are 99% internal UN papers of very limited interest to external audiences but I haven't had a professional reason to look at any EU agencies to confirm or deny value.
Additionally, the Financial Times has a second set of book awards along their main awards which you've not listed, the FT reader's best books list.
The article resonates, there is great serendipity in somewhat random exploration of books. But then the web app seems to be the antithesis of this? I clicked a few categories and the top ranked books all had the “popularity contest” smell, sort of the antithesis of random exploration.
The history section of the book price index is incredibly america-biased. History of america, american natives, japan after ww2 and vietnam in the last century.
That is what Sells. The USA has a lot of people so selling to us is going to make money even in a niche subject like history.
Historians who are not trying to sell mass market books have a much broader set of books. You won't see them in a normal bookstore.
And the irony of it is that the U.S. is actually quite history-poor, when compared to Europe, North Africa, Middle-East, Persia, China, Central Asia and India.
Can I make a feature request? I would love to be able to search by author. I recently started Caro's LBJ series (it's amazing so far) and I'm curious how many awards he's won. Anywho, I love non-fiction so I plan to visit this website often. Thank you and great work!
More generally, just sort by any of the column headers (year, title, publisher, author, score, ...) by clicking on it. In some views it works, in some it doesn't. A bit unintuitive.
The toggle between "traditional" text and "semantic" search is cool and new, but maybe the toggle could stay visible (it's there on the home tab, but then not anymore on the books tab).
At least one of the links on the post is still to localhost:3000 instead of the vercel page. I don't want to make a substack account to tell the author so hopefully this information finds its way over there.
Thanks for this! I've been looking for a good book on the Spanish Civil War, in either Spanish or English. The one American book I was recommended by someone, I didn't like. I'd like to keep searching regardless of the book's language or what country it has won prizes in. Could you also include book awards from other non-english speaking countries?
The war by itself is a far less interesting affair than how it happened, and what made the country be OK with fighting. For that you need to start at the end of the previous dictatorship, and cover the entire 2nd republic. Then the war starts making sense.
So I'd read The Collapse of the Spanish Republic, by Stanley Payne. He was a journalist in Spain through the entire period, so most of what he does is provide a first hand account. He never claims to try to be impartial, but he's pretty nouanced, as he knew a lot of the players personally.
Books on the Spanish Civil War I have read and can wholeheartedly recommend:
- "The Spanish Civil War" by Paul Preston
- "Homage to Catalonia" by George Orwell
Books on the same topic on my reading list (and quite frankly can't wait to get hold of them):
- "The Spanish Civil War" by Hugh Thomas
- "The Fight for Spain" by Antony Beevor
I've only read Homage to Catalonia out of those, which is a personal retelling from Orwell, are the other books in the same style? What's the difference you'd say between the ones you list?
Paul Preston's "The Spanish Civil War" would be the poster child of the "quality non-fiction" category specified in this post's title. He is, in my opinion, the leading Hispanist historian [1], and I hold his work in the highest esteem. I can also very much recommend his book "A People Betrayed".
Regarding the other two authors in my previous comment: Hugh Thomas was also a leading Hispanist historian [2], and Antony Beevor is a prominent military historian [3]. I expect both books to be in the same vein as Preston's.
(Quick aside for the astute reader that noticed all of the recommended authors being non-Spanish: I find Spanish politics extremely polarized; the Spanish Civil War being perhaps the most polarizing topic. As such, I find outsider's perspectives to be more balanced and objective.)
References:
1. https://en.wikipedia.org/wiki/Paul_Preston
2. https://en.wikipedia.org/wiki/Hugh_Thomas,_Baron_Thomas_of_S...
3. https://en.wikipedia.org/wiki/Antony_Beevor
> Quick aside for the astute reader that noticed all of the recommended authors being non-Spanish: I find Spanish politics extremely polarized; the Spanish Civil War being perhaps the most polarizing topic. As such, I find outsider's perspectives to be more balanced and objective
Makes sense to me, even as someone who live in Spain. It's a hard subject to talk to people about, especially because it never really was addressed as a country, everyone (politicians) just agreed to pretend it never happened, sweep a bunch of stuff under the rug and hope for the best. Then everyone act surprised when things start to bubble again...
I think the approach makes sense, also why you probably don't want to read about US politics from people actively inside of US politics, and it's similarly polarized today. Seemingly, at some point, people just lose all head and reason, and just start backing into their side, regardless of what that means.
+1 on this!
I very much have had the same experience as the author with research libraries. I have learned so many life-changing techniques and derivations from books that were two or three down from the book that I was looking for. When I became a research associate, I would often have my librarians ship the books to the branch closest to my office, and I stopped browsing the shelves myself. It wasn't until I went to the main library that I checked out a book several over from the one I had found in the catalog that I had a major breakthrough. Now that same library has collapsible shelving that makes browsing feel like a chore, and it makes me think we lost something valuable.
That said, I had high hopes for the catalog that the author had put together, but for my technical fields, I found it lacking. Trying to find books that would be close to van Soest's Nutritional Ecology of the Ruminant or van der Werf's Animal Breeding: Use of New Technologies, yields a whole bunch of pop-sci titles. The joy of finding these titles at the library in the past was how remarkably understandable and engaging they were while remaining scientifically rigorous. I immediately feel that the pop-sci titles that I was recommended are going to try to sell me something or some idea, rather than letting me explore ideas myself. I guess that means I need to go back to browsing library shelves myself.
On a related note, I was just noting to my co-founder, as we struggle to write good case studies for our website, that I find LLMs are astoundingly bad at writing good prose.
We all know the "AI-tics" that give away a sloppily AI-written piece, but even if you steer them, they still struggle to write consistently high-quality prose.
Somehow I feel that the work of a good copywriter has never been more noticeable.
I recently realized this as well and I think what I’ve discovered is that AI just produces mediocre content in all realms, but you don’t really notice it except in the realms where you have real expertise. With a lot of harness and prompting you can have it pump out something that’s pretty good but by default the next best token rarely produces anything of quality it seems like and if you think it does, perhaps you may want to recheck your assumptions on your expertise of the topic at hand
It tends to the mean. You can get it to do that less, but it's an inherent bias
I do think this explains much of it. But it doesn’t explain why Claude often writes in the clipped tone of that gnomic in-house tech evangelist who was inexplicably hired because he impressed the CEO.
(I may be projecting real life experience onto the LLM.)
I am genuinely fascinated as to how Claude acquired its utterly aggravating way of writing. It’s so much more irritating than ChatGPT, which is already not good.
Yeah the latest batch is really keyed on certain phrases - well beyond “you’re absolutely right!” Of the past. And the readability is trash almost always by default. Probably because they are forcing code so hard that it’s aligning prose into functional groups or something.
Interesting. So you think it’s somewhat emergent, in that sense, rather than a sort of designed tone of voice/writing style specified by supervised fine tuning (or the system prompt)?
I don’t see anyone having a Pulitzer Benchmark in their system card. ¯\_(ツ)_/¯
I've not really enjoyed finding out lately just how few people seem to notice what ought to be unmissable.
I think it's partly because of where they come from to the problem.
I have education/experience in both literature and coding, I have a pragmatic starting point when approaching text while also being able to recognize stylistic oddities, so I get to be the guy editing out AIsms sometimes.
But I've helped other people copywrite where their environment was all org-speak and academic writing, and AIsms don't really stand out in that case. AI is effectively "doing the right thing" writing the way it does for those tasks. Even tho the right thing is often a bad thing.
those people will also start adopting the AIsm and will become indistinguishable.
AI adopted humanisms, we just weren’t used to seeing them at the same scale we do today.
Diversity of writing styles was part of that, but I’d point to vernacular exposure as the larger component. We’re going to go through a period where we try to adapt to a form of “Universal English” for those of us who read primarily English writing.
Other languages’ readers may be experiencing the same dissonance when they come across AI-generated prose in their native language (but I’ll let others validate /reject my hypothesis).
It's like people learning a new language according to the book. Those people sound foreign to a native speaker. If AI was trained on school books for grammar, then they too will sound odd to a native speaker even if their output is technically fine or technically more proper. I known nothing of LLM training and how weighting is applied to educational content vs other sources, but it feels like they were weighted away from modern colloquial speak and towards book grammar. The whole thing reminds me of school literature classes where the teachers comments always felt like I was being guided to a more strained sound and less natural. But what do I know. Lit was my least favorite subject which is well evidenced by my grades compared to my math/science scores.
> we just weren’t used to seeing them at the same scale we do today.
I think there's also a lot of recency bias in it. The same thing that makes you suddenly notice how many people are driving the same model car you just looked at, or how many ads there are for Turbo Encabulators after you read an article about them. A lot of the "tells" people picked out in early AI are tells because they're also really common in the material that the AI was trained on and the styles it was made to emulate. But until everyone wanted "one quick trick" to pick out AI writings, people didn't have any particular reason to need to notice those tells and so they slipped under the radar.
This is right. AI is using human speech, but just doing so in a consistently peculiar way. The em-dash in particular is frustrating, because it's all over high-quality, pre-2022 academic work. But now instead of proper and erudite it's seen as AI-slop. Well, maybe if they didn't use it — all — time — ! It's even an auto-replacement in Word and can be (by one's choice) in LibreOffice, replacing three dashes (and two is replaced by an en-dash).
But when I submit a novel with em-dashes, will sloppy agents and sloppy editors be able to tell that em-dash was deliberately put there by me?
you're missing the point: the prevelance of AIsm language is spreading to people; I'm well aware it came from people.
All these HN posts with "stop with the LLM generated garbage" will slowly fade away because both people and LLMs will talk in a similar format and will be inured to the the distinguishing feature.
Certainly "humans" will keep carving out distinguishing characteristics, but just like "corporate speak" is a thing, AIsm will be a thing.
That's how language spreads.
Random observation: Google's Gemma 4 models write so much nicer prose than ChatGPT or Claude.
Though this might be me as a British reader, simply preferring a rather less American turn of phrase.
I reckon the more transatlantic, english-as-international language DeepMind team have had a subliminal (or maybe deliberate) impact on the way it chooses to write.
Or perhaps small open weights models simply aren't under the same commercial pressure to be engaging and sycophantic and are therefore less likely to adopt the samey overly casual, upbeat, Californian sales assistant manner. (Don't get me wrong, I like this from real human Californians just fine!)
Either way, the default tone is much less showy. I would be interested to find out if you agree.
I am very much an LLM cynic. I am engaging because I must, and trying to learn fundamentals, but I would not say I am overly excited by any of this, just glad that small open weights models exist as a counterpoint.
I loathe the way ChatGPT writes, and the Claude-isms that are everywhere; it is actually quite enraging, especially when you start seeing it in internet comments from people who used to try to write out their own thoughts.
But in my experiments with open weights models I have found I am much less aggravated by summaries and outlines written by Gemma 4, so much that I am happy enough to read them, because they have fewer irritants that take me out of the reading flow.
Though this evening it told me very kindly that my photography is a bit "safe". How very dare it… understand me that well.
[flagged]
[flagged]
[flagged]
You seem quite angry. I hope whatever this is passes and you feel better about yourself soon.
I gave Claude Fable $25 in Pangram API credits and, after hundreds of attempts, it was unable to produce a single readable original piece of writing that was not immediately identified as AI.
This seems to be a hard problem for LLMs, as passing would probably require good self-perception ("oh no, I am writing like an AI!") and fine-grained control over its own output ("let's write like a human instead!").
I think the actual way you'd do that is to run RLHF fine-tuning with Pangram as the judge.
I wonder if it partially because "write like a human" is kind of a vacuous request. Like, it's the objective everyone including me has been saying that we want, but there's no one way to write like a human and and humans don't even have a good definition past "I know it when I see it."
There's a lot of work in the humanities about different aspects of good writing, but that's not quite the same thing. And anyway they tend to assume a pre-existing level of writing ability. Students are supposed to learn good writing through practice; there are rules and exercises but they're incomplete.
I think as much it is that people write by grappling for the right phrase to represent some inner feeling or concept, writing in part for themselves, whereas LLMs write always and only for an audience.
It’s much easier to understand this once you think about other generative forms. MidJourney never just sits down and draws for fun, so fun never informs its art (only the outward appearance of others’ fun, separate from the fun itself). Suno doesn’t waste hours trying to find riffs on a guitar, so its output is never informed by the direct joy of getting it right. Its music is never optimised for playability on a particular guitar with a scratchy seventh fret and a too-high action. Neither Midjourney nor Suno have evolved their styles due to short-sightedness or carpal tunnel.
If you had a human writer who over a long career only ever wrote articles from an outline given to them by someone else, and you had all the outlines and all the resulting articles from those outlines, and you could train an LLM to generate an article from an outline, it still would not be kicking itself frustrated by an inelegant phrase in a prior article, it would not avoid certain phrases out of a passive aggressive reaction to some editor’s note, it would not ever just rush an article because everyone is gathering at the pub, and it would not choose an analogy just to rub the author of a bitchy critical letter to the editor the wrong way. An LLM could not “subtweet”. It could not write a series of articles hoping one important person will spot that they are auditioning for a job.
Creators have unseen, undocumented influences and motivations that inform their work over a long period. I don’t mean to say that these individual influences can be reliably detected in individual pieces of work. I do mean to say that I think their broad absence tends to be felt in LLM writing. As readers we develop an affinity for writers as much as for their writing, and we do this in part because we deduce things about them.
Very elegantly put! So much so in fact, that I finally put down my phone to grab my laptop in order to write this. I have had a similar much cruder version of this thought: LLMs are not human, they don't know what it's like to stub their toe, despite having read probably millions descriptions of it. Even amongst humans there are experiences which are impossible to share, despite us having evolved to communicate quite effectively. Writing is just a small part of it, and cannot replace the actual experience of being human.
[dead]
Each person writes in a different personal way, so writing “like a human” would actually require a model being able to purposefully make the specific choices that an individual human writer does.
However, general purpose LLMs like Fable have been trained on huge amounts of all kinds of data, and therefore find it exceedingly hard to break out of the grooves carved by that data. They can’t avoid defaulting to centroids and averages, even when they are trying not to. This makes it possible for classifiers like Pangram to discriminate their writing.
A plausible way to work around this limitation would be to train a LLM on a limited and cohesive subset of writing materials, so it would absorb their specific writing style.
One example might be Talkie, a LLM trained on pre-1930’s English text. Talkie is a far smaller and less powerful model than Fable.
And yet, Talkie’s writing is so distinctive that it is often classified as human by Pangram.
Is pangram using knowledge about specific models? Could you throw it off by having several models generate a paragraph each or so?
A lot of what would be the top "reference" works aren't even that good either, they were a successful marketing phenomenon or had cultural or social relevance at their time. So, you can get a lot of bad prose going by a number of somewhat logical, externally measurable parameters.
More noticeable to me is the lack of the work of a good copy editor which, sadly, we haven't had for a really long time. At least, not on the interwebs. Even the news sites reduced where their print copies were known for rigorous editing saw obvious issues with the various corporate overlords doing serious headcount reductions. The rush to be first to publish reduced even further the time any editors might have had, and then the wide spread use of CMS style articles that slammed output together with something as unintelligent as 'cat segmentFromAuthor1 segmentFromAuthor2 segmentFromAuthor3 > article' where you can tell where each segment started over again with the same basic information as if it was content meant to stand on its own.
Of course, the amount of self published work has also helped make the lack of a good copy editor noticeable. I can excuse self published blogs though. But the stuff released "professionally" has really become farcical.
I can sniff out AI writing immediately but from what I hear AI writing is more popular than ever
Worth stating that this is one of the success stories from AI. Someone who has domain expertise outside of programming is able to create a really useful piece of software because the barrier to entry has been significantly lowered. Really nice.
If you are interested in a non-fiction book, check out Veritas: Truth Across Cultures[1]. It's a collection of quotes/saying/proverbs that are similar between cultures, which I call truths. I am in interested in long term thinking, things that stood the test of time. This was primarily put together for myself as a lens to view the world.
[1] https://www.chestergrant.com/veritas
Very neat site and write-up, but I found this part amusing:
> There is really nothing “AI” about this aside from the tool that collected the data and coded it, and, crucially, semantic search […]
So really, everything about it is AI. And that’s not a bad thing! It’s okay to simultaneously preach the superiority of award winning books over AI-generated garbage while also acknowledging the same AI as a valuable tool for other uses.
Key word being tool. People get lost in a fantasy that the AI was responsible for decisions and not the human who chose to use it as a replacement for effort rather than a tool to enhance effort, then apply their bias to all AI use.
so much of AI is like, to me, used as was to generate basically dynamic applicaions. if we ever got anywhere with semantics, objects, etc, pushed html further, maybe we'd naturally have web clients that could take objects, natively search, collate, display, put on a map, graph, etc... but AI is like a text only shortcut to the end result we want.
I have a stupid AI app I use now for some bookkeeping which was previously, just an excel sheet. except it has a UI, it validates inputs, and does a bunch of other housekeeping stuff that makes it nice. Did the excel work? yes totally.
Penguin Random House seems to be dominating the leaderboard. What's the story here? Are there only very few non-fiction publishers capable of publishing high-quality work, or is Penguin Random House exceptionally good at it?
Fascinated that the author shares an experience of discovery in libraries. The internet, in its early days felt like that. Of course, no longer.
Libraries are certainly declining in their traditional form. I find it odd that everyone has a digital resources in their pocket yet libraries are squeezing out physical books to make way for more and more computers. Try to find paper copy of the Sony founder's book.... that will be £50 on amazon. Prohibitive. No library within a 10 mile radius has a copy.
I remember joyfully discovering Tony Royce's book in my library a few years ago. That enlightenment will never happen now. Primary knowledge is being lost, churned crude will forever lubricate the delusions of those who have no facility to collate the basis of our understanding.
> The internet, in its early days felt like that. Of course, no longer.
back in the reasonably early days of the web (I would guess ~2000) I stumbled across a webpage that said "I used to collect random interesting snippets in a shoebox, here is the internet version of it". I was delighted enough that I actually wrote to the author to compliment them on helping keep the web interesting, and though I haven't thought about it in ages it clearly made enough of an impression on me that I remembered the author's name. and it's still up! https://www-users.york.ac.uk/~ss44/cyc/index.htm
AIs are fantastic for discovery. Just today I asked Chatgpt to find blogs 1) about old school newspaper cartoons 2) that have been writing for more than 10 years and 3) are run by one or two passionate people rather than a team. Within seconds Chatgpt found 7 candidates. Within another handful of seconds it gave me RSS links to the ones I wanted.
This is incredible!
Sometimes. I think this is partly because search engines are so poor nowadays. I have asked AI for certain websites that I know exist but it can't find them.
This hits the nail on the head regarding the real utility of AI right now. The magic isn’t in using LLMs to generate endless reams of synthetic text; it’s in using embeddings and semantic search to navigate high-signal, human-curated data.
My bona-fides here amount to little more than being a big time reader of non-fiction. But IMO, there's a nice synthesis here. Rather than going long rounds of asking LLMs about a subject, I usually end up asking it for book recommendations. It's much better than a google search and you can push it into some deep corners if you go past the surface level recs.
You can get quite specific. You can find texts you wouldn't discover unless you spent years studying the topic. Often these are completely approachable and give interesting perspectives they just get buried behind a wall of syllabi and listicles.
This is how I ended up reading Thompson's 'The Making of the English Working Class' and Graves' 'Goodbye to all That' among others.
I did this for a recent trip! I asked books related to or authored in the places I was visiting. Found some interesting titles.
Is it even possible to find one these days? I would love to see some real honest amazon kindle stats on the amount of ebooks added over the last 3 years. Even a honest pie chart to show people its not worth it to get them to stop posting them.
"To avoid the tedium, I decided that I would also flip to a random page of every book I shelved and read a random sentence from it."
Also, this is about the lost joy of "browsing".
I wonder how much serendipity played when wandering through a library (or other things in life—even window shopping). But now, everything on the internet, you more or less have to know what it is you're looking for—are rarely "surprised or delighted".
I’m not sure I agree with that.
I’ve found plenty of interesting things by randomly wandering around on Gutenberg.org, completely undirected.
If they have a sort of "wander UI", that's cool. I'm not sure what that looks like though.
This is wonderful. 30 seconds into trying the site and I’ve already found a couple of interesting books to read. The search I typed is ‘Stuff being made, how it was made, how history shaped its form’.
When I see lists like this I am surprised that they always live out school textbooks.
School textbooks are the kind of books that have the very precise purpose of giving you a foundation and roadmap to explore and understand a whole field of knowledge. And they're an industry that has been improving them on successive iterations to produce highly organized, comprehensive and accessible reading material through years.
School textbooks also have a lot of bias, and may oversimplify issues in ways that are necessary with children, but lead to longing misconceptions.
IE: Was the US Civil War fought over slavery or states' rights? Someone's answer may come from the biases of the textbooks they grew up with.
> School textbooks also have
For me the keyword in your comment is "also". What this word means to me is that, whatever faults textbooks might have, there is no alternative solution. They replicate a problem that all books have.
However, within the domain of textbooks, you still might find good solutions for these bias. These bias are mostly within history and social sciences and good and reputable Universities are renowned for publishing books in these domains that are quite balanced, comprehensive, deep and rigorous. Oxford and Cambridge are the 2 best examples that come to my mind. The U.S. has good Universities too, but after their government declared a war on science and knowledge I'd be carefull before trusting these universities.
I would really like this, but using academic reading lists instead of prizes. What are the important historians of (e.g.) medieval Italy?
Tangentially related, but what a snappy webpage
This is pretty nice, I might use this if I'm looking to learn more on a topic.
I'm always cautious when reading nonfiction because it's hard for me to tell if the author knows what they're talking about when I'm not an expert myself. Using awards is a good metric. Maybe.
I am curious what the "score" for each book means. Is it calculated by giving each award a certain weight and adding them up?
Thank you! Wonderful idea, and very well executed. Useful discovery tool.
I understand your passion is non-fiction, but would be wonderful to have the equivalent for fiction, of which there is a great deal of excellent literature. Perhaps the same site could incorporate both fiction and non-fiction, with a filter?
maybe I am being too small brained black/white about this, but something feels amiss about the idea of vibecoding a solution to an AI generated problem. How can we get better, doing the same thing that made us sick?
I love the "books for dads who like Pavement" prompt! It would be great if you could incorporate the same experience of "browsing" random page of some of the results that are returned
I like very very much that there's no user voting system here.
I just recently read a really mediocre book with a 4.7/5 rating on Goodreads. Shows being sloppy is not exclusively a privilege for neither the AIs or producers.
I read at night in bed. I love how, with ebooks (first Overdrive, now Libby), when I hear about an interesting book, I can within 30 seconds search for it across multiple libraries, borrow it, and send it to my Kindle.
Thanks for making this, I searched up economics and found at least 3 titles that , when reading through first pages, are really really interesting.
Is it only me or the titles of top books from History, Nature, and Politics sound very click-baiting? I'm really taken off by these titles.
Hey buddy, are you feeling all right?
Awesome stuff!
Now, can this interface with Libby? The worldcat interface is, well, confusing. I can't figure out if it knows libraries near me exist.
I've gotten AI to write things that made me teary eyed --- it manipulated me
I made a sandwich yesterday and cried for an hour because it reminded me of the way my mom used to make school lunches. No AI necessary.
Serious question: can we define the top 5 parameters that distinguish human-generated non-fiction writing from AI "slop" writing (I don't like the term "slop" because it carries a built-in bias)?
Since AI output is based on the most likely next word/token given the context, let's consider how humans do this, i.e. choose the next word based on context.
Is one of the parameters the degree of "randomization"?
Are these other parameters?
- Specific experience?
- Recent exposure to relevant materials, i.e. paraphrasing?
- Borrowed knowledge from a specific resource?
- Writing style, i.e. choice of words or sentence structure?
I'd like to hear what others think.
I think your amazon links are broken, the ones i tried all 404.
Could you please add a link to the Goodreads page of each book?
Thank you for sharing. This will come in handy for my research!
I'm not sure how I feel about "Entangled Life" on the first page of books in "science". Isn't that slop? I mean, it's pop-science bullshit, and it's based on feelgood extrapolation from a thin evidence base. Maybe that doesn't quite fit the definition of slop, but it's not a signal of quality for the rest of the list.
Literally reading the books like an LLM would be trained on them!
As it happens, I had the same job as the author at uni. I shelved books at the biomedical and sciences library. As a life-long book lover, it was a dream uni job being surrounded by all sorts of historical treatises on medical arcana.
I've been really exercising my non-fiction reading muscles of late. I'm trying weaponise it as an antidote to AI slop and doom-scrolling. My last read was Designing Data-Intensive Applications by Martin Klepmann (the newly released second edition), it was a wonderful read!
The antithesis of AI slop is when you've created a unique idea from first principles and AI keeps misunderstanding your idea and spell checking it back to something that's widely popular such that you have to write a whole extra alignment subsystem to check that it hasn't ignored half of what you've wrote in your first principles document and just thought you were talking about something distantly related in its training set.
To work on something truly original in non-fiction in the future against AI pushing everything to the mean of its training set and fighting against that when doing research, as opposed to generating a whole book of slop from familiar ideas, is going to be quite the unrecognized effort of writing those books.
But I get triggered by the em dashes. :(
Fiction genre catching strays as usual
possibly a bug - I searched for "writers like lewis thomas" and the top two results were by thomas himself.
On a similar note: on Steam, the original Dark Souls game[1] is listed with the "souls-like" tag. That's not wrong, I guess, but just kind of a tautology. Grice's maxims[2] generally have such tautologies omitted in ordinary conversation as non-informative.
[1]: https://store.steampowered.com/app/570940/DARK_SOULS_REMASTE...
[2]: https://en.wikipedia.org/wiki/Cooperative_principle
For me it is surprisngly natural to include the original thing into "a set of thing-likes".
Why does it feel wrong (or meaningless) to you?
I really thought ChatGPT would be able to tell me what this is called but it can't. I suggest "circular eponym".
It's like calling a CRT monitor a really big flatscreen.
it's more like you would call a xerox-brand photocopier a "xerox machine" and mean that it was a photocopier rather than that it was made by xerox.
For embeddings based search I think this is unavoidable
And what pray tell is everyone doing with their privileged leisure time?
You have access to every single wicked problem in the world and you choose to do nothing.
Shame
I participate book club for several years with friends, we casually navigate certain topics, and i can tell you human slop in literature is a real thing.
- almost every book try to stretch core idea into book size format
- unique ideas are rare people attack them under different angels
- a later phenomena : their believes almost predict entire book, outcomes etc, brainwash impact is real
so not sure, how ai slop is better vs book slop, at least with ai you can distill the idea, with the book, you have to spend 10-40 hours to digest average, absolutely non fresh ideas, that author brought in just to sell that book, otherwise it would be magazine article worth.
This is sadly correct/true. This is one of the many reasons why I militantly reject carbon chauvinism. Most of the potshots people take against AI can be easily retorted with “have you seen how bad the average human is at X?”
> how ai slop is better vs book slop, at least with ai you can distill the idea
Why don't you just use whatever LLM you have on hand to summarize every book you could read instead of reading it? The idea that ai slop is 'better' because you can distill the idea is an insane point to me because it treats books as a pure consumption-based concept, where the goal is strictly to finish a book and move on.
I think even the worst book is more valuable than whatever garbage an LLM writes because you can at least learn from the intent. Where and why the author failed, where they got stuck, the points they failed to make. LLM writing is a void, there's nothing to be learned.
In your experience, how do you determine book slop? For me I don't really stray from authors and topics I like, and non-fiction isn't really my go to outside of sci-fi, and mystery/law books. Like on the stretch point, I can tell that in other media but not so much books.
Business books are 99% slop.
Self improvement books! All of them! "Make good habits, not bad habits because this is what Lincoln did..."
They’ve used a pretty good filter for the human slop - prestigious awards.
Sweet, I think I found a few to read
> And before you wonder, yes this is actually free. I am paying for the hosting and the API costs entirely because I just want people to find and read more good non-fiction books.
Just to note, you do have affiliate links so it isn't 100% out of the goodness of your heart ;) Not blaming you or anything, just saying...
I think the antithesis of ai slop is meditation on a nice beach. Or eating like a really good meal. Or holding your child.
Not to be that guy but your site is terrible.
It does not do the most important function well, giving me a list of award winning books that is easy to sort through. It is not clear what is a link and what is not.
It looks fine though somewhat generic. But it is not actually functional in a real sense.
real nice work!
I don't think we would have AI slop, if they were just trained on high quality media.
Less relatable AI, maybe, but not sloppy AI.
AI trained on other AI (i.e. distilling), can be better than the AI they were trained on, likely because of this reason. Each iteration can raise the bar.
As apposed to the deep spiral of generating shiny attention crystals for humans, and then training on that.
The slop isn't a property of the LLM but the workflow it's utilized in. Just now I asked ChatGPT about the linguistic effect where a CRT could be described as a giant flatscreen, a landline could be described as a wired cellphone, or the original Rogue could be described as a roguelike. Unfortunately it doesn't know what that's called either, but that is a non-slop use of an LLM, like a semantic search engine. On the other hand if you print LLM output into a book it's almost certainly slop no matter how the model was trained.
Training AI on AI will most likely increase slop. LLM's add randomness in their outputs, books usually don't.
That isn't correct.
Obviously, you can handicap any model with an indiscriminate dataset. Are you training on slop? Why?
More data is not automatically better data.
Just three (of many) advantages of quality data selection: (1) It gives a clearer signal. (2) It, ironically, reduces the complexity of what needs to be learned, since slop adds its own complexity. (3) It reduces dataset size which means more learning per watt (and per just about any other cost). The advantages compound.
Humans are no different. Children surprise with their ability to absorb sophisticated relationships and skills, while people who had bad examples struggle to recover.
99% of human effective intelligence is higher quality cultural knowledge learned as children. None of us had to spend centuries deciding whether zero, negative numbers, the square root of -1 are numbers. Slop + quality is not quality.
why read when you can flick on the tube?
It's sad now to see a university library where students sit among endless shelves of amazing books, sitting on laptops, almost all of them with ChatGPT open. The library has become just a place to sit and open an LLM.
It's especially sad because on of the things those LLMs are best, almost purpose built for is to tell those students which books to go open, where to find nuggets people haven't bumped into for years, to make cross-connections that would take a PhD a decade to find.
This is a very good point. If it should be used as anything, it is as a pointer to other things.
The most sad part of all is to think how little of that knowledge is digitized and available to be referenced by ChatGPT. Digitizing old books is one of the highest things you can do as a human for “good for the world”.
The title is tautological. Quality is the opposite of slop, AI or not.
Not really. Quality nonfiction books are a subset of quality books.
the cope from authors.
Nice site, but most of the top biographies are US figures: Edgar Hoover, Frederick Douglass, Lyndon Johnson etc. I'd like to see more non-US biographies.
[dead]
[flagged]
[dead]
[flagged]
[flagged]
[dead]
[dead]
[flagged]
Can you please review the site guidelines (https://news.ycombinator.com/newsguidelines.html) and stick to them when posting? You broke several of them here.
"Don't be snarky."
"Omit internet tropes." (the stopped-reading-at bit)
"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."
Edit: it looks like your account has been posting quite a few flamebait and/or unsubstantive comments generally. Could you please not do that? It's not what this site is for, and destroys what it is for.
> Decries AI for producing slop
> Uses AI to produce website
The article isn't even about AI slop other than the title for some reason
It's a clever title because it's FREE MONEY on HN. There is a small but rabid group that upvotes everything that denigrates AI, sight unseen.
I think my issue with this project--and so many other similar ones--is that the provenance of the code does undermine the intention. If a project purports to be about quality, then knowing that the creator abdicated some of the responsibility for creating the thing they ostensibly care about makes it harder to put faith in them as having high standards elsewhere.
Perhaps I am just old-fashioned, and vibe-coding is something that can be done with full focus and care for high quality, but I remain unconvinced. This is a project that needs to be cared about sincerely to be trustable/meaningful/useful, and the approach taken casts doubt on that.