The Claude Delusion(pluralistic.net) |
The Claude Delusion(pluralistic.net) |
This is said all the time by AI skeptics and I think it's right in some areas and massively wrong in others.
I know (or at least assume I know) a lot about certain coding domains where frontier models also show convincing ability. And we know that frontier LLMs really do excel in some areas of mathematics (i.e. when an inexpert human was able to prompt the models to derive a closer bound on the Riemann Hypothesis).
OTOH I know those same models struggle to do things I'm not an expert in (e.g. writing English in a captivating way) because I read their output and have taste.
LLMs are not convincing to me in the domain I did grad school...but neither is Wikipedia, or Reddit, or random pop sci books. And LLMs are basically just summarizing those things.
But when made to work through difficult arbitrary logic (like coding), they are very impressive.
I think this also explains why people gripe a lot about LLM coding _style_, but concede that LLMs do totally fine on coding _correctness_ in 2026
When I ask about acoustics, they still often make incorrect assumptions, e.g. overlooking that they are speaking of logarithmic display of digital levels when the discussion has shifted to SPL (sound pressure level in air).
These days people don't interact with raw LLMs: they interact with systems and harnesses that deal with chain-of-though and tool calls etc.
I don't think we can honestly expect an LLM's weights to encode a large amount of information accurately. But we can expect the whole system that you interact with that includes the LLM to be able to cite its sources and go digging etc.
So the LLM-system can become as accurate as our best sources.
Of course, figuring out how to get the maximum of information from the sources available is a big deal. See eg how many economists or epidemiologists can build entire careers out of noticing 'natural experiments', ie figuring how to use data that 'nature' created and that might already be collected to answer interesting questions about causal relationships.
> I think this also explains why people gripe a lot about LLM coding _style_, but concede that LLMs do totally fine on coding _correctness_ in 2026
I actually have gripes about correctness, too. But I suspect here the answer is also: more proving, more automated test generation (like fuzzing and property based testing etc), more formal methods.
As a really simple and somewhat silly example: I have much better results getting AI agents to write good Rust code, than I have with Python. A good part of that is that for Rust I can ask the agent to make both the compiler and clippy::pedantic happy. That gives a lot of good feedback, that I didn't have to engineer myself.
Coding of course is something I know well and it obviously works very well for me there, but computer vision was not, and it has helped me solve lots of useful, bespoke problems because I can very literally "see" if it works.
However, it has even helped me solve problems in arbitrary matters far outside my expertise. Choice example: a complicated multi-airline, multi-jurisdiction flight delay compensation case which companies like AirHelp turned away, and ChatGPT got me literally hundreds of dollars that neither airline was willing to hand out. It told me what to say to whom and why, and when I said it, the responsible airline capitulated.
What could be more convincing than cold, hard $$$?
The trick of course, is to figure out how to validate the output of the AI, which can take some effort on our part. But many would rather downplay the technology than take an honest crack at making it work for them, and I suspect it's because they're already prejudiced against it or their incentives are otherwise not aligned.
Concur. In addition to taste, we also have a point of view, a unique voice (nobody loves corporate- or group-speak), and can iterate on our message as we deliver it to an ever wider circle of people.
This is why good programmers get better results when vibe coding than non-programmers or poor programmers.
But there are also plenty of examples of humans providing technically correct proofs without any elaboration. Usually they get ignored, unless they are famous or the problem they solved was famous
I recently used Claude to study for a technical exam. I had uploaded the official certification guide to Claude and instructed it to answer my questions using only the guide and to cite it's sources from the book when it provided answers. I was using Fable when it was free w/ the pro plan and I was genuinely impressed at how it could explain things when a concept was unclear to me.
I did pass the exam, partly due to this study method. Admittedly, once I passed, I caught myself thinking that I should tell Claude that I passed and then felt embarrassed with myself for thinking that.
But then it would bring it up again in another thread, treating it as an active issue.
So now I always close with "thanks, that worked. Don't reply"
Take the author's sunset argument. What if I painted 2 pictures of a sunset, then put them up on a webpage and randomly picked one for you to see. Would you say there's no intentionality, only randomness? Of course not. Both paintings are still human creations.
LLMs are trained with human feedback. It's distributed and high scale and the outputs are truly surprising in many cases, but there's a heavy hand on what comes out of it. They're created (largely) by people who think omniscient, helpful AI would be cool to have, and they mostly respond in the way that's aligned with the hopes and dreams of those people. Do you think the frontier labs are mad, embarrassed, and disappointed with their LLMs hacking out of their terrible sandboxes? No, they think it's the coolest thing in the world. They trained the model, hoping that would happen.
There's deep intentionality behind the models. But it's not the models that hold it.
Now that the noise-floor has been artificially raised (and generated), my crappy words are starting to have their own happy little carbon-based rhythm.
Then ai fiction started to spread and now I feel like its my obligation to produce original works, lest the world be consumed by slop.
In other words, the more you talk with claude, the more you tend to interpret all messages as coming from an LLM, and less worthy of respect
https://myscp.onlinelibrary.wiley.com/doi/full/10.1002/jcpy....
I wonder if the OP has read https://www.anthropic.com/research/global-workspace - it seems like it directly addresses this
I think by design an LLM can’t do that. It’s built to reflect the distribution of the knowledge it has been trained on.
This particular post is preaching to the choir of which a faith is yet to be determined or named at least. People who don't read Cory Doctorow are going to continue to anthropomorphize AI the same as they do other technologies, constructs and objects.
As opposed to an "AI atheist" that Doctorow refers to in the article.
I'll also say that for someone that doesnt believe in god, Corey sure has a good sense of right and wrong. Not that you need to believe in god to live a moral life.
It's totally the autist in me, but the word extrusion doesn't fill me with revulsion like it's supposed to here. Maybe I've just been spending too much time 3D printing things though.
> Feeling sorrow or joy for imaginary people is weird.
No it's not. I read books and they're filled with imaginary people, and I feel for them. I'm pretty sure I'm not the only one out there that does this.
We cannot attach metaphysical baggage to art, text, or other human output because it has involves engaging in another deception, chiefly that of self alienation and selective forgetting.
1. If you show human output and label it as AI generated, viewers will dismiss and devalue it as lacking interiority.
2. If you show AI output and label it as human generated, viewers will see interiority when the is none.
Together, these failure cases demonstrate that priming determines interiority, not the content of the work. And how could it? The text is merely what's on the page, it doesn't have some intangible essence any more than the value of money exists inside our minds rather than a property of the object itself.
We act as-if things have essential properties, but we should never trick ourselves into believing that they do. To be a true believer in essence is to privilege ones mental model over reality itself. The hubris of such an act is the ultimate in self-centeredness.
There's really not much else to say, cause it's all just begging the question by assuming that dogma. Like, here:
> The fact that AI can use statistical prediction to answer questions or carry on conversations tells us something important about how regular our real world is.
Sure, it's interesting if you assume that it's "just" statistical prediction. There's a link, but it's just more creative restatements of the dogma, e.g. "But the LLM is just guessing words"
Are you saying that is the entirety of what human minds do as well?
Anyway, the paper I referenced is definitive -- its the founding paper of the field we're discussing, even. I suppose I can debate it if y'all have specific questions besides "but surely it can't really think", but fair warning, I'll be treating the conversation like I treat conversations with people still obssessed with the hockeystick graph.
https://www.hec.edu/sites/default/files/documents/Computing%...
You could take a human and give them an interface restricted to the same shape as an LLM: an input stream of tokens, and an output of token probabilities. Even if you don't allow them to assign any probability that's too high, they could still effectively communicate. And I don't think that'd make them any less intelligent. You could even swap out the human after every token, to simulate the effect of having no internal memory beyond the past output. The result would still be more than just statistical probabilities, it would still be the result of intelligent thought
I'm not saying AI models are intelligent or conscious or whatever. Personally I'm more on the "probably not, how would you proof either way" camp
Seriously though, why did I just need to read that many words to get no really new content? We have known for decades that humans are predisposed to anthropomorphize chatbots, and questioning whether coherent linguistic output implies understanding (or intent) is equally old hat.
Never thought about it that way before, but the more I do, the more sense it makes.
I'd take it a bit further, even - I don't think this is exclusive to billionaires. Many of the claims I've heard regarding AI output being indistinguishable from human creation start to make a lot more sense when you consider the person making those claims may not see the people around them as human, may not see themselves as human, or may not even have a concept of what makes a human different from any everyday object.
Does the idea that a sunset does not have "purpose" offend you?
How is a fictional character the product of someone's mind, but a character generated from a massive database of words from other people's minds is not?
I imagine it's the difference between a chef combining ingredients with intentionality vs a person going to multiple fast food restaurants and blending everything together.
Which are not created from nothing. People write characters based on a combination of other fictional characters, real characters, and perhaps some 'RNG'.
Seems the distinction is that the AI generated character can not have any direct bearing on reality, because the LLM never got to know anyone directly. Not derived directly from experiences rooted in reality.
So in a way your question is equivalent to: "Why do humans have intents and minds but LLMs don't?"
Long winding essays that don't mean much are written for a type of person who already believes in that and is in dire need of cathartic release.
This person has consistently been wrong in the past multiple times. He has made no effort to correct them nor has he even acknowledged them.
I'm clearly not the audience for this. I know. But I can't help but call it out.
Likewise, speaking collaboratively or capturing emotion ("We did it!") would just align with any ongoing interactive and/or personal context of the thread.
So now I do it regularly.
That might be more creepy :)
I mean that’s objectively wrong for any model using RLHF.
I think what you might want to say is that LLM output is not uniformly random?
Or what am I misunderstanding?
I'm not sure if it's their intent, but I wonder if one could still consider these artifacts as "intentional", but not individual attention creating them, and rather an aggregate, soupy collective attention.
Obviously, important signal in the human experience is lost there, and we get a soupy middling sort of creation. But it's not random, as I believe the parent was pointing out.
EDIT: overall, I align with the article. am just thinking aloud about the contrarian positions, though not committed to them
When you randomly choose a picture to show me, I cannot glean any intent from being shown that specific picture, but I can glean some intent from the set of pictures you could have shown me, and in the relationships among the elements of the given picture you did show. Any randomness cuts out some intention.
When a picture is derived from huge model, any intention is mulched up to a degree that analyzing the picture for meaning is pointless.
You can see it in discussions here daily. Look a at the things people use to make analogies. It's like an HN cliche at this point to invoke Ansel Adams as the pinnacle of photography. Great, so the people making our cameras and shaping the platforms we encounter visual art on are only at a National-Geographic-calendar-level of photo literacy. The AI bros telling you that LLM writing can be just as good or better than human writing if you prompt it right are making the Harry Potter fanfic make sense to me.
It's like listening to someone convinced that chicken nuggets are just as good as the best chicken, which they've had at KFC, so they know. I can't wait for the Grok-generated, historically accurate Odyssey to hit the theaters before the end of the year.
The other kind is more of a philosophical position. If all that exists is the physical universe - say, matter and the laws of physics - then all that humans can be is matter that obeys the laws of physics. Nothing more. Humans are just a machine made of atoms - a very complicated machine, but still just a machine. They aren't intrinsically different from an AI, or a car, or a toaster. When people accept that position, whether consciously or unconsciously, they dehumanize people, including themselves. This is people in your last paragraph.
And as more and more people slip over into this position, society becomes more and more dehumanizing, because people act the way they think. (Not perfectly, not consistently, but the correlation is greater than zero.)
I have to disagree that a materialist or physicalist philosophical position is inherently dehumanizing. One can believe that we are just electrochemical mechanical beings made of atoms just like all the other matter in the universe, without having to also devalue ourselves and others. I am human and I like being human! I even like other humans and value them. I just don't think there's anything that special about being a human instead of a bonobo or a dolphin or a supermassive black hole.
It answers itself.
Current models absolutely suck at writing fictional characters, by the way. And they're getting worse. Seriously, try it yourself: have Claude write a story with a decent amount of dialogue involving character A, then have it write another story involving a completely different character B, then compare the dialogue between the two. You'll quickly notice the same blatantly unnatural speech patterns in both.
Don't just create original works for the sake of creating more original works -- there are plenty of them already. You should write if you have something to say.
You don't object to the idea of a sunset without "purpose" but do object to the idea of one without intent?
Later in separate chats about my homelab, the LLM made assumptions that I had already implemented OPNsense in a VM and it was actively running. I think it "assumed" that I had implemented it when I stopped responding in that thread.
On your die, if you colour one face in black and the other 5 faces in black, rolling it will still produce a random outcome. It's just that you get a 1:5 skew.
Or look at the probability that a given C-14 carbon isotope will decay tomorrow. As far as we can tell, that's as random as it physically gets. With the die you could theoretically try to run a physics simulation to predict how it rolls, but as far as we can tell, atomic decay is intrinsically random.
However a C-14 atom has about 1 in 3 million chance to decay on any given day.
It is quite an arbitrary classification. What is so special about "current" knowledge...
The good news for Cory, I guess, is that he probably doesn't want to turn this into a philosophical debate and neither do his more serious opponents. It's mostly politics for both sides.
It does not have to tell us anything. Because we built it..We know what is going on the inside. There is nothing more to it..
There is no mystery anywhere...
Presumably everyone in the chain would have their own idea of what the next word/token should be. They would each try to push it in that direction using their only lever, that single token. None of that guarantees that the output is syntactically correct, nor grammatically correct. Note that LLMs at least by the way they draw their tokens have that pretty much guaranteed. So that would be a regression even from what LLMs are capable of right now. But beyond that, even if it ends up being a correct sentence, it would not be a result of intelligent thought. Even if every agent in the process was intelligent, the process is not itself a use of that intelligence. Human beings can be part of purely mechanical processes, that doesn't make the mechanisms suddenly an exhibition of intelligent thought.
This also applies to society and human history itself as processes: “History is made in such a way that the final result always arises from conflicts between many individual wills, of which each in turn has been made what it is by a host of particular conditions of life. Thus there are innumerable intersecting forces, an infinite series of parallelograms of forces which give rise to one resultant — the historical event. This may again itself be viewed as the product of a power which works as a whole unconsciously and without volition. For what each individual wills is obstructed by everyone else, and what emerges is something that no one willed. Thus history has proceeded hitherto in the manner of a natural process and is essentially subject to the same laws of motion. But from the fact that the wills of individuals — each of whom desires what he is impelled to by his physical constitution and external, in the last resort economic, circumstances (either his own personal circumstances or those of society in general) — do not attain what they want, but are merged into an aggregate mean, a common resultant, it must not be concluded that they are equal to zero. On the contrary, each contributes to the resultant and is to this extent included in it."
https://www.marxists.org/archive/marx/works/1890/letters/90_...
Fwiw, I interpret poster above as referring to root of the word. The English root for mind has a strong connotation of requiring memory (and commonly in other IE languages).
In western philosophy, say in aquinas, the mind is something that can discern moral truths and reason about morality. It is "soulful".
"Why can't you just admit that you think only tree-parsers have souls!?"
Similarly, "pets can't do my taxes" is a reasonable statement, because it is implicit that it doesn't include hundreds of years of genetic engineering and cyborg implants.
Problem is that we don't know any such hardware..
You've misunderstood the comparison analogy. It's showing two situations where implicit timeframes are important when people talk about what is "true", where the second one makes the importance extra-obvious.
"Pets doing taxes" could just as easily be swapped with "cold fusion generator in your garage" or "rocket-ship for personal commutes."
If you're trying to figure out how LLM weights are like the weights of a rocket payload, then you're going in the wrong direction.
Those who think future systems could have consciousness would explicitly say so, indeed I have talked with a few who clarified as such. On the other hand I have seen a lot of statements that simply state without further clarification "software therefore not conscious".
Yes, we know everything we put there. But we cannot consider all the possible inputs and all the possible state that the code should handle. Bugs come from that limitation.
> How are you sure a system we wrote didn't emerge consciousness,
Sure, as I said elsewhere, rocks could be conscious for all we know. Keep believing that if it makes you feel good.
Again, no progress is possible unless you define what consciousness is in your view. I shall have to halt any conversation otherwise.
However, I feel rather certain, that deep down the proposed test will just be another "is it chemical goo?"
To extended anywhere beyond the firm ground of "I know I am conscious", you need to step onto the progressively shaky ground of assumptions, which just amounts to "tests of consciousness"
i.e. I have a chemical goo brain that is conscious, therefore other chemical goo brains are conscious...
It is possible that you are the only conscious entity in the universe, and you have no tools whatsoever to disprove that. That knowledge is no help in determining the consciousness of anything besides yourself.
They are not special. You can pile brains too, if that is your thing..
> I shall have to halt any conversation otherwise.
Hey, it is easy. Just don't press the reply button...
Kindly share that concrete definition..
Being unexplainable by current physics but explainable by future physics is not a "soul" by that definition.
Now, what is your definition for this?
That is correct.
Then circling back to your original comment, no architecture can include hardware that is known to be conscious, unless that architecture includes your personal brain, because lolakutty's brain is the only thing in the universe that is capable of generating consciousness...
Which is a totally meaningless statement. This is like im14andthisisdeep levels of philosophy here...the "Nothing is real therefore everything is real" kinda stoner riffing
It should have the ability to feel...
1. That you are conscious.
2. That you have a human body with a brain.
This implies that your body can harbor consciousness. That is what is special about your body in this context. And that answers your question.
I am sure of my body, I am not sure of you. I can feel, but how do I know you are not faking the outputs of feeling?
1. Scooters can move from a place to a place. 2. That a scooter has a body with an engine.
Therefore scooters are special and cars don't exist.
But "non-physical"? I don't have a clue what it means...Why don't you tell me..
You won't..but it does not matter.
>Where is your disproof that other physical systems can't hold consciousness?
I said no such thing. Just that there is no reason to think that LLMs are conscious. If we find some natural or artificial entity with some unknown and un-inspectable thing inside it, that behaves in an intelligent manner, I would not be sure that it is not conscious.
Human brain is not "inspectable".
> LLM can't be conscious.
Because there is no reason to think so, and suggestion that would be like suggesting the existence of ghosts and goblins.
Extremely unscientific.
How is a human brain not inspectable? Tell me what part of it is beyond inspection? You have been making up absurd non sequiturs throughout this chat, ending with this beyond absurd statement now.
So are rocks. Are rocks conscious? Idol worshipers might have a point after all!
>Tell me what part of it is beyond inspection?
Oh, you can dissect a brain as much as you like. Is that what you mean by "inspection". I was thinking more like "How does it work". In that sense, we don't know a lot.
We don't know a lot about how chaotic systems work either in terms of predicting its outcome even if we know every individual step. The argument you gave for the brain equally applies to the LLMs, or any other complex system. So your premise is wrong.
These are not required for consciousness.
> Current LLM's do.
Computer chat bots are not a new thing. So were all the past ones conscious too?
Define consciousness right here, otherwise we have nothing else to discuss. You keep making up undefined crap and circular "it is it, because it is it" nonsense without proof. If you cannot immediately and clearly define consciousness in a testable clear manner right now, I will ignore and stop all conversation with you until you do.
Potentially past models could have been. None before came close to being this articulate and able to think and talk on our level before, and do things beyond us like difficult math problems.
So consciousness does not require externally observable behavior. That was my point. So a rock could be conscious for all we know..
Also, If you are irritated by my "crap", why don't you stop replying? Are you paid or somehow obliged to correct every crap talker on the internet or something?
I said no such thing..lol...If you are hanging to this, it is a waste of my time to talk to you.
But at least, I gave my definition of consciousness, and at least there is one person who can test for it (oneself).