Funny, I've used them to create my own personalized text editor, perfectly tailored to what I actually want. I'm pretty sure that didn't exist before.
It's wild to me how many people who talk about LLM apparently haven't learned how to use them for even very basic tasks like this! No wonder you think they're not that powerful, if you don't even know basic stuff like this. You really owe it to yourself to try them out.
I've worked at multiple AI startups in lead AI Engineering roles, both working on deploying user facing LLM products and working on the research end of LLMs. I've done collaborative projects and demos with a pretty wide range of big names in this space (but don't want to doxx myself too aggressively), have had my LLM work cited on HN multiple times, have LLM based github projects with hundreds of stars, appeared on a few podcasts talking about AI etc.
This gets to the point I was making. I'm starting to realize that part of the disconnect between my opinions on the state of the field and others is that many people haven't really been paying much attention.
I can see if recent LLMs are your first intro to the state of the field, it must feel incredible.
So it is absurdly incorrect to say "they can only reproduce the past." Only someone who hasn't been paying attention (as you put it) would say such a thing.
Like, I'm sorry, but you're just flat-out wrong and I've got the proof sitting on my hard drive. I use this supposedly impossible program daily.
This is even more clear in the case of diffusion models (which I personally love using, and have spent a lot of time researching). All of the "new" images created by even the most advanced diffusion models are fundamentally remixing past information. This is really obvious to anyone who has played around with these extensively because they really can't produce truly novel concepts. New concepts can be added by things like fine-tuning or use of LoRAs, but fundamentally you're still just remixing the past.
LLMs are always doing some form of interpolation between different points in the past. Yes they can create a "new" SQL query, but it's just remixing from the SQL queries that have existed prior. This still makes them very useful because a lot of engineering work, including writing a custom text editor, involve remixing existing engineering work. If you could have stack-overflowed your way to an answer in the past, an LLM will be much superior. In fact, the phrase "CRUD" largely exists to point out that most webapps are fundamentally the same.
A great example of this limitation in practice is the work that Terry Tao is doing with LLMs. One of the largest challenges in automated theorem proving is translating human proofs into the language of a theorem prover (often Lean these days). The challenge is that there is not very much Lean code currently available to LLMs (especially with the necessary context of the accompanying NL proof), so they struggle to correctly translate. Most of the research in this area is around improving LLM's representation of the mapping from human proofs to Lean proofs (btw, I personally feel like LLMs do have a reasonably good chance of providing major improvements in the space of formal theorem proving, in conjunction with languages like Lean, because the translation process is the biggest blocker to progress).
When you say:
> So it is absurdly incorrect to say "they can only reproduce the past."
It's pretty clear you don't have a solid background in generative models, because this is fundamentally what they do: model an existing probability distribution and draw samples from that. LLMs are doing this for a massive amount of human text, which is why they do produce some impressive and useful results, but this is also a fundamental limitation.
But a world where we used LLMs for the majority of work, would be a world with no fundamental breakthroughs. If you've read The Three Body Problem, it's very much like living in the world where scientific progress is impeded by sophons. In that world there is still some progress (especially with abundant energy), but it remains fundamentally and deeply limited.
If a programmer creating their own software (or contracting it out to a developer) would be a bespoke suit and using software someone or some company created without your input is an off the rack suit, I'd liken these sorts of programs as semi-bespoke, or made to measure.
"LLMs are literally technology that can only reproduce the past" feels like an odd statement. I think the point they're going for is that it's not thinking and so it's not going to produce new ideas like a human would? But literally no technology does that. That is all derived from some human beings being particularly clever.
LLMs are tools. They can enable a human to create new things because they are interfacing with a human to facilitate it. It's merging the functional knowledge and vision of a person and translating it into something else.
Put another way, and I hate to throw in the now over-used phrase, but I feel you may be responding to a strawman that doesn't much appear in the article or the discussion here: "Because these tools don't achieve a god-like level of novel perfection that no one is really promising here, I dismiss all this sorta crap."
Especially when I think you are also admitting that the technology is a fairly useful tool on its own merits - a stance which I believe represents the bulk of the feelings that supporters of the tech here on HN are describing.
I apologize if you feel I am putting unrepresentative words in your mouth, but this is the reading I am taking away from your comments.
Also , a shitton of what we do economically is reproducing the past with slight tweaks and improvements. We all do very repetitive things and these tools cut the time / personnel needed by a significant factor.
From what you've described an LLM has not invented anything. LLMs that can reason have a bit more slight of hand but they're not coming up with new ideas outside of the bounds of what a lot of words have encompassed in both fiction and non.
Good for you that you've got a fun token of code that's what you've always wanted, I guess. But this type of fantasy take on LLMs seems to be more and more prevalent as of late. A lot of people defending LLMs as if they're owed something because they've built something or maybe people are getting more and more attached to them from the conversational angle. I'm not sure, but I've run across more people in 2025 that are way too far in the deep end of personifying their relationships with LLMs.
That is a derived output. That isn't new as in: novel. It may be unique but it is derived from training data. LLMs legitimately cannot think and thus they cannot create in that way.
Curious, does it perform at the limit of the hardware? Was it programmed in a tools language (like C++, Rust, C, etc.) or in a web tech?
Do you only take LLM seriously if it can be another Einstein?
> But a world where we used LLMs for the majority of work, would be a world with no fundamental breakthroughs.
What do you consider recent fundamental breakthroughs?
Even if you are right, human can continue to work on hard problems while letting LLM handle the majority of derivative work
For a more practical example, creating bindings from dynamic-language-A for a library in compiled-language-B is a genuinely useful task, allowing you to create things that didn't exist before. Those things are likely to unlock great happiness and/or productivity, even if they are derived from training data.
Also the derived argument doesn’t really hold, just because you know about two things doesn’t mean you’d be able to come up with the third, it’s actually very hard most of the time and requires you to not do next token prediction.
I for one think your work is pretty cool - even though I haven't seen it, using something you built everyday is a claim not many can make!
The change hit us so fast a huge number of people don’t understand how capable it is yet.
Also it certainly doesn’t help that it still hallucinates. One mistake and it’s enough to set someone against LLMs. You really need to push through that hallucinations are just the weak part of the process to see the value.
Why is that kind of thinking required to create novel works?
Randomness can create novelty.
Mistakes can be novel.
There are many ways to create novelty.
Also I think you might not know how LLMs are trained to code. Pre-training gives them some idea of the syntax etc but that only gets you to fancy autocomplete.
Modern LLMs are heavily trained using reinforcement data which is custom task the labs pay people to do (or by distilling another LLM which has had the process performed on it).
You don’t have a solid background. No one does. We fundamentally don’t understand LLMs, this is an industry and academic opinion. Sure there are high level perspectives and analogies we can apply to LLMs and machine learning in general like probability distributions, curve fitting or interpolations… but those explanations are so high level that they can essentially be applied to humans as well. At a lower level we cannot describe what’s going on. We have no idea how to reconstruct the logic of how an LLM arrived at a specific output from a specific input.
It is impossible to have any sort of deterministic function, process or anything produce new information from old information. This limitation is fundamental to logic and math and thus it will limit human output as well.
You can combine information you can transform information you can lose information. But producing new information from old information from deterministic intelligence is fundamentally impossible in reality and therefore fundamentally impossible for LLMs and humans. But note the keyword: “deterministic”
New information can literally only arise through stochastic processes. That’s all you have in reality. We know it’s stochastic because determinism vs. stochasticism are literally your only two viable options. You have a bunch of inputs, the outputs derived from it are either purely deterministic transformations or if you want some new stuff from the input you must apply randomness. That’s it.
That’s essentially what creativity is. There is literally no other logical way to generate “new information”. Purely random is never really useful so “useful information” arrives only after it is filtered and we use past information to filter the stochastic output and “select” something that’s not wildly random. We also only use randomness to perturb the output a little bit so it’s not too crazy.
In the end it’s this selection process and stochastic process combined that forms creativity. We know this is a general aspect of how creativity works because there’s literally no other way to do it.
LLMs do have stochastic aspects to them so we know for a fact it is generating new things and not just drawing on the past. We know it can fit our definition of “creative” and we can literally see it be creative in front of your eyes.
You’re ignoring what you see with your eyes and drawing your conclusions from a model of LLMs that isn’t fully accurate. Or you’re not fully tying the mechanisms of how LLMs work with what creativity or generating new data from past data is in actuality.
The fundamental limitation with LLMs is not that it can’t create new things. It’s that the context window is too small to create new things beyond that. Whatever it can create it is limited to the possibilities within that window and that sets a limitation on creativity.
What you see happening with LEAN can also be an issue with the context window being too small. If we have an LLM with a giant context window bigger than anything before… and pass it all the necessary data to “learn” and be “trained” on lean it can likely start to produce new theorems without literally being “trained”.
Actually I wouldn’t call this a “fundamental” problem. More fundamental is the aspect of hallucinations. The fact that LLMs produce new information from past information in the WRONG way. Literally making up bullshit out of thin air. It’s the opposite problem of what you’re describing. These things are too creative and making up too much stuff.
We have hints that LLMs know the difference between hallucinations and reality but coaxing it to communicate that differentiation to us is limited.
I don't believe they can. LLMs have no concept of truth.
What's likely is that the "truth" for many subjects is represented way more than fiction and when there is objective truth it's consistently represented in similar way. On the other hand there are many variations of "fiction" for the same subject.
5 years ago a typical argument against AGI was that computers would never be able to think because "real thinking" involved mastery of language which was something clearly beyond what computers would ever be able to do. The implication was that there was some magic sauce that human brains had that couldn't be replicated in silicon (by us). That 'facility with language' argument has clearly fallen apart over the last 3 years and been replaced with what appears to be a different magic sauce comprised of the phrases 'not really thinking' and the whole 'just repeating what it's heard/parrot' argument.
I don't think LLM's think or will reach AGI through scaling and I'm skeptical we're particularly close to AGI in any form. But I feel like it's a matter of incremental steps. There isn't some magic chasm that needs to be crossed. When we get there I think we will look back and see that 'legitimately thinking' wasn't anything magic. We'll look at AGI and instead of saying "isn't it amazing computers can do this" we'll say "wow, was that all there is to thinking like a human".
Mastery of words is thinking? In that line of argument then computers have been able to think for decades.
Humans don't think only in words. Our context, memory and thoughts are processed and occur in ways we don't understand, still.
There's a lot of great information out there describing this [0][1]. Continuing to believe these tools are thinking, however, is dangerous. I'd gather it has something to do with logic: you can't see the process and it's non-deterministic so it feels like thinking. ELIZA tricked people. LLMs are no different.
[0] https://archive.is/FM4y8 [0] https://www.theverge.com/ai-artificial-intelligence/827820/l... [1] https://www.raspberrypi.org/blog/secondary-school-maths-show...
This is the definition of a derived product. Call it a derivative work if we're being pedantic and, regardless, is not any level of proof that LLMs "think".
This is the definition of the word ‘novel’.
You’re using ‘derived’ to imply ‘therefore equivalent.’ That’s a category error. A cookbook is derived from food culture. Does an LLM taste food? Can it think about how good that cookie tastes?
A flight simulator is derived from aerodynamics - yet it doesn’t fly.
Likewise, text that resembles reasoning isn’t the same thing as a system that has beliefs, intentions, or understanding. Humans do. LLMs don't.
Also... Ask an LLM what's the difference between a human brain and an LLM. If an LLM could "think" it wouldn't give you the answer it just did.
Back to the land of reality... Describing something in fiction doesn’t magically make it "not an invention". Fiction can anticipate an idea, but invention is about producing a working, testable implementation and usually involves novel technical methods. "Star Trek did it" is at most prior art for the concept, not a blueprint for the mechanism. If you can't understand that differential then maybe go ask an LLM.
So think of it like this, to create the model we use terabytes of data. Then we do RL which is probably less than one percent of additional data involved in the initial training.
The change in the model is that reliability is increased and hallucinations are reduced at a far greater rate than one percent. So much so that modern models can be used for agentic tasks.
How can less than one percent of reinforcement training get the model to tell the truth greater than one percent of the time?
The answer is obvious. It ALREADY knew the truth. There’s no other logical way to explain this. The LLM in its original state just predicts text but it doesn’t care about truth or the kind of answer you want. With a little bit of reinforcement it suddenly does much better.
It’s not a perfect process and reinforcement learning often causes the model to be deceptive an not necessarily tell the truth but it more gives an answer that may seem like the truth or an answer that the trainer wants to hear. In general though we can measurably see a difference in truthfulness and reliability to an extent far greater than the data involved in training and that is logical proof it knows the difference.
Additionally while I say it knows the truth already this is likely more of a blurry line. Even humans don’t fully know the truth so my claim here is that an LLM knows the truth to a certain extent. It can be wildly off for certain things but in general it knows and this “knowing” has to be coaxed out of the model through RL.
Keep in mind the LLM is just auto trained on reams and reams of data. That training is massive. Reinforcement training is done on a human basis. A human must rate the answers so it is significantly less.
I imagine that sounded more profound when you wrote it than it did just now, when I read it. Can you be a little more specific, with regard to what features you would expect to differ between LLM and human responses to such a question?
Right now, LLM system prompts are strongly geared towards not claiming that they are humans or simulations of humans. If your point is that a hypothetical "thinking" LLM would claim to be a human, that could certainly be arranged with an appropriate system prompt. You wouldn't know whether you were talking to an LLM or a human -- just as you don't now -- but nothing would be proved either way. That's ultimately why the Turing test is a poor metric.
That's the crazy thing. Yes, in fact, it turns out that language encodes and embodies reasoning. All you have to do is pile up enough of it in a high-dimensional space, use gradient descent to model its original structure, and add some feedback in the form of RL. At that point, reasoning is just a database problem, which we currently attack with attention.
No one had the faintest clue. Even now, many people not only don't understand what just happened, but they don't think anything happened at all.
ELIZA, ROFL. How'd ELIZA do at the IMO last year?
Either that, or they tried it "last year" or "a while back" and have no concept of how far things have gone in the meantime.
It's like they wandered into a machine shop, cut off a finger or two, and concluded that their grandpa's hammer and hacksaw were all anyone ever needed.
After post-training, this is definitively NOT what an LLM does.
Language is the substrate of reason. It doesn't need to be spoken or written, but it's a necessary and (as it turns out) sufficient component of thought.
Without you, there was nothing.
I can think of several offhand.
1. The effect was never real, you've just convinced yourself it is because you want it to be, ie you Clever Hans'd yourself.
2. The effect is an artifact of how you measure "truth" and disappears outside that context ("It can be wildly off for certain things")
3. The effect was completely fabricated and is the result of fraud.
If you want to convince me that "I threatened a statistical model with a stick and it somehow got more accurate, therefore it's both intelligent and lying" is true, I need a lot less breathless overcredulity and a lot more "I have actively tried to disprove this result, here's what I found"
SWEs are trained to discard surface-level observations and be adversarial. You can't just look at the happy path, how does the system behave for edge cases? Where does it break down and how? What are the failure modes?
The actual analogy to a machine shop would be to look at whether the machines were adequate for their use case, the building had enough reliable power to run and if there were any safety issues.
It's easy to Clever Hans yourself and get snowed by what looks like sophisticated effort or flat out bullshit. I had to gently tell a junior engineer that just because the marketing claims something will work a certain way, that doesn't mean it will.
First, the claim that RLHF materially reduces hallucinations and increases factual accuracy is not anecdotal. It shows up quantitatively in benchmarks designed to measure this exact thing, such as TruthfulQA, Natural Questions, and fact verification datasets like FEVER. Base models and RL-tuned models share the same architecture and almost identical weights, yet the RL-tuned versions score substantially higher. These benchmarks are external to the reward model and can be run independently.
Second, the reinforcement signal itself does not contain factual information. This is a property of how RLHF works. Human raters provide preference comparisons or scores, and the reward model outputs a single scalar. There are no facts, explanations, or world models being injected. From an information perspective, this signal has extremely low bandwidth compared to pretraining.
Third, the scale difference is documented by every group that has published training details. Pretraining consumes trillions of tokens. RLHF uses on the order of tens or hundreds of thousands of human judgments. Even generous estimates put it well under one percent of the total training signal. This is not controversial.
Fourth, the improvement generalizes beyond the reward distribution. RL-tuned models perform better on prompts, domains, and benchmarks that were not part of the preference data and are evaluated automatically rather than by humans. If this were a Clever Hans effect or evaluator bias, performance would collapse when the reward model is not in the loop. It does not.
Fifth, the gains are not confined to a single definition of “truth.” They appear simultaneously in question answering accuracy, contradiction detection, multi-step reasoning, tool use success, and agent task completion rates. These are different evaluation mechanisms. The only common factor is that the model must internally distinguish correct from incorrect world states.
Finally, reinforcement learning cannot plausibly inject new factual structure at scale. This follows from gradient dynamics. RLHF biases which internal activations are favored, it does not have the capacity to encode millions of correlated facts about the world when the signal itself contains none of that information. This is why the literature consistently frames RLHF as behavior shaping or alignment, not knowledge acquisition.
Given those facts, the conclusion is not rhetorical. If a tiny, low-bandwidth, non-factual signal produces large, general improvements in factual reliability, then the information enabling those improvements must already exist in the pretrained model. Reinforcement learning is selecting among latent representations, not creating them.
You can object to calling this “knowing the truth,” but that’s a semantic move, not a substantive one. A system that internally represents distinctions that reliably track true versus false statements across domains, and can be biased to express those distinctions more consistently, functionally encodes truth.
Your three alternatives don’t survive contact with this. Clever Hans fails because the effect generalizes. Measurement artifact fails because multiple independent metrics move together. Fraud fails because these results are reproduced across competing labs, companies, and open-source implementations.
If you think this is still wrong, the next step isn’t skepticism in the abstract. It’s to name a concrete alternative mechanism that is compatible with the documented training process and observed generalization. Without that, the position you’re defending isn’t cautious, it’s incoherent.
The key point you’re missing is the type of failure. Search systems fail by not retrieving. Parrots fail by repeating. LLMs fail by producing internally coherent but factually wrong world models. That failure mode only exists if the system is actually modeling and reasoning, imperfectly. You don’t get that behavior from lookup or regurgitation.
This shows up concretely in how errors scale. Ambiguity and multi-step inference increase hallucinations. Scaffolding, tools, and verification loops reduce them. Step-by-step reasoning helps. Grounding helps. None of that makes sense for a glorified Google search.
Hallucinations are a real weakness, but they’re not evidence of absence of capability. They’re evidence of an incomplete reasoning system operating without sufficient constraints. Engineers don’t dismiss CNC machines because they crash bits. They map the envelope and design around it. That’s what’s happening here.
Being skeptical of reliability in specific use cases is reasonable. Concluding from those failure modes that this is just Clever Hans is not adversarial engineering. It’s stopping one layer too early.
And beyond people claiming that LLMs are basically sentient you have people like CamperBob2 who made this wild claim:
"""There's no such thing as people without language, except for infants and those who are so mentally incapacitated that the answer is self-evidently "No, they cannot."
Language is the substrate of reason. It doesn't need to be spoken or written, but it's a necessary and (as it turns out) sufficient component of thought."""
Let that sink. They literally think that there's no such thing as people without language. Talk about a wild and ignorant take on life in general!
From the abstract: "Is thought possible without language? Individuals with global aphasia, who have almost no ability to understand or produce language, provide a powerful opportunity to find out. Astonishingly, despite their near-total loss of language, these individuals are nonetheless able to add and subtract, solve logic problems, think about another person’s thoughts, appreciate music, and successfully navigate their environments. Further, neuroimaging studies show that healthy adults strongly engage the brain’s language areas when they understand a sentence, but not when they perform other nonlinguistic tasks like arithmetic, storing information in working memory, inhibiting prepotent responses, or listening to music. Taken together, these two complementary lines of evidence provide a clear answer to the classic question: many aspects of thought engage distinct brain regions from, and do not depend on, language."
What's funny is the failure to grasp any contextual framing of ELIZA. When it came out people were impressed by it's reasoning, it's responses. And in your line of defense it could think because it had mastery of words!
But fast forward the current timeline 30 years. You will have been of the same camp that argued on behalf of ELIZA when the rest of the world was asking, confusingly: how did people think ChatGPT could think?
The mental gymnastics here is entertainment at best. Of course the thinking LLM would give feedback on how it's actually just a pattern model over text - well, we shouldn't believe that! The LLM was trained to lie about it's true capabilities in your own admission?
How about these...
What observable capability would you expect from "true cognitive thought" that a next-token predictor couldn’t fake?
Where are the system’s goals coming from—does it originate them, or only reflect the user/prompt?
How does it know when it’s wrong without an external verifier? If the training data says X and the answer is Y - how will it ever know it was wrong and reach the correct conclusion?
https://pmc.ncbi.nlm.nih.gov/articles/PMC2799957/
The resources that the brain is using to think -- whatever resources those are -- are language-based. Otherwise there would be no way to communicate with the test subjects. "Language" doesn't just imply written and spoken text, as these researchers seem to assume.
You need to read a few papers with publication dates after 2023.
Meanwhile, you didn't answer my question. How'd ELIZA do on the IMO? If you know a way to achieve gold-medal performance at top-level math and programming competitions without thinking, I for one am all ears.
Its annoying to see posts from people who lag behind in intelligence and just dont get it - people learn at different rates. Some see way further ahead.
If you want to go around huffing and puffing your chest about a subject area, you kinda do fella. Credibility.
This is the internet bro. Credibility is irrelevant because identities can never be verified. So the only thing that matters is the strength and rationality of an argument.
That’s the point of hacker news substantive content not some battle of comparison of credentials or useless quips (like yours) with zero substance. Say something worth reading if you have anything to say at all, otherwise nobody cares.
If you want a financial tip, don't short stock and chase market butterflies. Instead, make real professional friends, develop real skills and learn to be friendly and useful.
I made my money in tech already, partially by being lucky and in the right place at the right time, and partially because I made my own luck by having friends who passed the opportunity along.
Hope that helps!
Absolutely not true. I cannot express how strongly this is not true, haha. The tech is neat, and plenty of real computer scientists work on it. That doesn't mean it's not wildly misunderstood by others.
> Concluding from those failure modes that this is just Clever Hans is not adversarial engineering.
I feel like you're maybe misunderstanding what I mean when I refer to Clever Hans. The Clever Hans story is not about the horse. It's about the people.
A lot of people -- including his owner-- were legitimately convinced that a horse could do math, because look, literally anyone can ask the horse questions and it answers them correctly. What more proof do you need? It's obvious he can do math.
Except of course it's not true lol. Horses are smart critters, but they absolutely cannot do arithmetic no matter how much you train them.
The relevant lesson here is it's very easy to convince yourself you saw something you 100% did not see. (It's why magic shows are fun.)
These things are not horses. How can anyone choose to remain so ignorant in the face of irrefutable evidence that they're wrong?
https://arxiv.org/abs/2507.15855
It's as if a disease like COVID swept through the population, and every human's IQ dropped 10 to 15 points while our machines grew smarter to an even larger degree.
He doesn't care. You might as well be arguing with a Scientologist.
How can you think without tokens of some sort? That's half of the question that has to be answered by the linguists. The other half is that if language isn't necessary for reasoning, what is?
We now know that a conceptually-simple machine absolutely can reason with nothing but language as inputs for pretraining and subsequent reinforcement. We didn't know that before. The linguists (and the fMRI soothsayers) predicted none of this.
In other words, when I (seem to) dismiss an entire field of study, it's because it doesn't work, not because it does work and I just don't like the results.
Funnily enough, they did, if you go back far enough. It's only the deconstructionists and the solipsists who had the audacity to think otherwise.
That's the exact reason I mention the Clever Hans story. You think it's obvious because you can't come up with any other explanation, therefore there can't be another explanation and the horse must be able to do math. And if I can't come up with an explanation, well that just proves it, right? Those are the only two options, obviously.
Except no, all it means is you're the limiting factor. This isn't science 101 but maybe science 201?
My current hypothesis is the IMO thing gets trotted out mostly by people who aren't strong at math. They find the math inexplicable, therefore it's impressive, therefore machine thinky.
When you actually look hard at what's claimed in these papers -- and I've done this for a number of these self-published things -- the evidence frequently does not support the conclusions. Have you actually read the paper, or are you just waving it around?
At any rate, I'm not shocked that an LLM can cobble together what looks like a reasonable proof for some things sometimes, especially for the IMO which is not novel math and has a range of question difficulties. Proofs are pretty code-like and math itself is just a language for concisely expressing ideas.
Here, let me call a shot -- I bet this paper says LLMs fuck up on proofs like they fuck up on code. It will sometimes generate things that are fine, but it'll frequently generate things that are just irrational garbage.
The first thing I checked was "how did they verify the proofs were correct" and the answer was they got other AI people to check it, and those people said there were serious problems with the paper's methodology and it would not be a gold medal.
https://x.com/j_dekoninck/status/1947587647616004583
This is why we do not take things at face value.
Clever Hans was exposed because the effect disappeared under controlled conditions. Blind the observers, remove human cues, and the behavior vanished. The entire lesson of Clever Hans is not “people can fool themselves,” it’s “remove the hidden channel and see if the effect survives.” That test is exactly what has been done here, repeatedly.
LLM capability does not disappear when you remove human feedback. It does not disappear under automatic evaluation. It does not disappear across domains, prompts, or tasks the model was never trained or rewarded on. In fact, many of the strongest demonstrations people point to are ones where no human is in the loop at all: program synthesis benchmarks, math solvers, code execution tasks, multi-step planning with tool APIs, compiler error fixing, protocol following. These are not magic tricks performed for an audience. They are mechanically checkable outcomes.
Your framing quietly swaps “some people misunderstand the tech” for “therefore the tech itself is misunderstood in kind.” That’s a rhetorical move, not an argument. Yes, lots of people are confused. That has no bearing on whether the system internally models structure or just parrots. The horse didn’t suddenly keep solving arithmetic when the cues were removed. These systems do.
The “it’s about the people” point also cuts the wrong way. In Clever Hans, experts were convinced until adversarial controls were applied. With LLMs, the more adversarial the evaluation gets, the clearer the internal structure becomes. The failure modes sharpen. You start seeing confidence calibration errors, missing constraints, reasoning depth limits, and brittleness under distribution shift. Those are not illusions created by observers. They’re properties of the system under stress.
You’re also glossing over a key asymmetry. Hans never generalized. He didn’t get better at new tasks with minor scaffolding. He didn’t improve when the problem was decomposed. He didn’t degrade gracefully as difficulty increased. LLMs do all of these things, and in ways that correlate with architectural changes and training regimes. That’s not how self-deception looks. That’s how systems with internal representations behave.
I’ll be blunt but polite here: invoking Clever Hans at this stage is not adversarial rigor, it’s a reflex. It’s what you reach for when something feels too capable to be comfortable but you don’t have a concrete failure mechanism to point at. Engineers don’t stop at “people can be fooled.” They ask “what happens when I remove the channel that could be doing the fooling?” That experiment has already been run.
If your claim is “LLMs are unreliable for certain classes of problems,” that’s true and boring. If your claim is “this is all an illusion caused by human pattern-matching,” then you need to explain why the illusion survives automated checks, blind evaluation, distribution shift, and tool-mediated execution. Until then, the Hans analogy isn’t skeptical. It’s nostalgic.
Your fallback move here is rhetorical, not scientific: “maybe it doesn’t mean what you think it means.” Fine. Then name the mechanism. What specific process produces internally consistent multi-step proofs, respects formal constraints, generalizes across problem types, and fails in ways analogous to human reasoning errors, without representing the underlying structure? “People are impressed because they’re bad at math” is not a mechanism, it’s a tell.
Also, the “math is just a language” line cuts the wrong way. Yes, math is symbolic and code-like. That’s precisely why it’s such a strong test. Code-like domains have exact semantics. They are adversarial to bullshit. That’s why hallucinations show up so clearly there. The fact that LLMs sometimes succeed and sometimes fail is evidence of partial competence, not illusion. A parrot does not occasionally write correct code or proofs under distribution shift. It never does.
You keep asserting that others are being fooled, but you haven’t produced what science actually requires: an alternative explanation that accounts for the full observed behavior and survives tighter controls. Clever Hans had one. Stage magic has one. LLMs, so far, do not.
Skepticism is healthy. But repeating “you’re the limiting factor” while refusing to specify a falsifiable counter-hypothesis is not adversarial engineering. It’s just armchair disbelief dressed up as rigor. And engineers, as you surely know, eventually have to ship something more concrete than that.
The question wasn’t “are humans sometimes self-delusional?” Everyone agrees with that. The question was whether, in this specific case, the prevailing view about LLM capability is meaningfully wrong in a way that has implications. If you really believed this was mostly Clever Hans, there would be concrete consequences. Entire categories of investment, hiring, and product strategy would be mispriced.
Instead you retreated to “don’t short stocks” and generic career advice. That’s not skepticism, it’s risk-free agnosticism. You get to sound wise without committing to any falsifiable position.
Also, “I made my money already” doesn’t strengthen the argument. It sidesteps it. Being right once, or being lucky in a good cycle, doesn’t confer epistemic authority about a new technology. If anything, the whole point of contrarian insight is that it forces uncomfortable bets or at least uncomfortable predictions.
Engineers don’t evaluate systems by vibes or by motivational aphorisms. They ask: if this hypothesis is true, what would we expect to see? What would fail? What would be overhyped? What would not scale? You haven’t named any of that. You’ve just asserted that people fool themselves and stopped there.
Gemini 2.5 has since been superceded by 3.0, which is less likely to need hints. 2.5 was not as strong as the contemporary GPT model, but 3.0 with Pro Thinking mode enabled is up there with the best.
Finally, saying, "Well, they were given some hints" is like me saying, "LOL, big deal, I could drag a Tour peleton up Col du Galibier if I were on the same drugs Lance was using."
No, in fact I could do no such thing, drugs or no drugs. Similarly, a model that can't legitimately reason will not be able to solve these types of problems, even if given hints.
I've spent a lot of time feeding similar problems to various models to understand what they can and cannot do well at various stages of development. Reading papers is great, but by the time a paper comes out in this field, it's often obsolete. Witness how much mileage the ludds still get out of the METR study, which was conducted with a now-ancient Claude 3.x model that wasn't at the top of the field when it was new.
Here, let me call a shot -- I bet this paper says LLMs fuck up on proofs like they fuck up on code. It will sometimes generate things that are fine, but it'll frequently generate things that are just irrational garbage.
And the goalposts have now been moved to a dark corner of the parking garage down the street from the stadium. "This brand-new technology doesn't deliver infallible, godlike results out of the box, so it must just be fooling people." Or in equestrian parlance, "This talking horse told me to short NVDA. What a scam."