Excellent optimistic post in a sea of negativity, and with actual suggestions, too.
After reading, my mental image is this: think of Olympiads in Ancient Greece.
* A weightlifter was only awarded a laureate if he were able to lift a heavy stone (have no idea what they were lifting, for illustrative purposes only :-)
* Along comes Archimedes who invents what we would call an exoskeleton. Now any regular guy can lift twice as much as last year’s athlete.
* What to do? You can cancel the Olympiads, but they are actually useful as training, motivation, etc So now you have to give the prize on other factors, eg how well he can lift, has he opened a gym in the city, etc
BTW, physics and bio are not exempt, so those researchers better read and try to stay ahead.
I don´t think so. For programming agents can run code, check compiler output, etc. For mathematics, it is almost the same once you factor in the usage of lean.
For the reality, you can´t close the loop that fast, or with that precision. You will have to slow down by several orders of magnitude.
Doesn’t have to. There are petabytes of experimental physics data that can be fed to AI to extract additional insights. The only holdup is this is slightly harder to than math.
With bio, you’re right, generally designing and conducting an experiment goes hand in hand and theoretical biologist is not a common label.
Unfortunately, the totality of the evidence very much indicates that Sanborn went "buck wild" with the enciphering, he made mistake(s), or both. So this is very much in line with the Chaocipher challenge of 1990. Nice little earner for some people though.
Thought-provoking article. When the cave painters were at work, they most probably didn’t see themselves as “artists” and of course not scientists, philosophers. That was an age where all those things were together at once, living in a religious state, I think. Little by little, they separated (to sometimes unhappy results). Brings to mind the old myth where humans were both male and female originally, but gods, fearing their power, separated them, leading us to look for our half. Same with art and science, I think.
So while Picasso is perhaps right to say “We have invented nothing “ as quoted in the article when he means art is as old as humans, the realization that something is art, is perhaps invented afterwards.
I also reject the two kinds of art thesis that’s proposed here:
“It follows that there must also be two kinds of art – similarly different and yet, somehow, complementary. The first is what I call artifice: art conceived so as to confirm, reinforce and extend the empire of facts in which we ordinarily live. … Examples come readily to mind: propaganda, pornography and advertising; sentimental, moralistic or didactic art”
Many current artists would lump, say, paintings by Old Masters into this category. Facing the statue of David past month in Florence, it was eye opening how much such a classic statue of a human can move people!
? Sentimental means excessively moving, so how is that showing that the statue of David doesn't fit the category (actually here my quibble is with the article, not you), and, just because people miscategorize it in your opinion, how does that mean there's no such category?
Edit: I see the article is laying claim to being properly, meaningfully moved. It seems that by including "sentimental" the article is saying: Through art, we are in touch with the truth and beauty of the mystery of creation itself, but you are just having an unnecessary moment because you saw a cliche.
I happen to be re-reading Deleuze right now (and knew Bergson at some point), and I'm not really sure what the author adds to the equation. I.e. art is the pure desire-machine, schizophrenic and completely deterritorialized, whereas pornography is the desire-machine territorialized by determinate conditions of e.g. OnlyFans. If anyone else knows Deleuze I'd be curious; him and Guattari is not exactly a vibe hang. Long-winded way of saying it seems like a lazy dichotomy.
My data point is I got to stare at Wheatfield with Crows and The Death of Marat within 24 hours and I was walking around Brussels a drooling mess, only capable of leaving decoy cheap sunglasses for pickpockets
"All art is, in the end, adoration" J.Campbell. (quoting by memory), essentially art and humanity are one in those times, no one invented the other.
In more recent times things are changed, life and art are on different path, but still in reinessance times people received education bot in liberal arts and in scientific (by the times) matters. Art and science still talks under the surface.
Oh boy, this reminded me of The Boy Who Sailed Around the World Alone by Robin Lee Graham, which I randomly found and read in my school's library when I was 12 or 13. Still remember it 40+ yrs later.
I too loved this book, though my copy is titled _Dove_ - one of my all time favorites. My dad, who gave me the book, knew Robin during his own solo sailing days, and figured I could relate to a lot of what he wrote about. Sadly I never got a chance to meet him myself though.
Similar to how Cypher puts it: I know this is “just” next token inference, matrix mult and just software, ie there’s no “intelligence” there BUT, looking at this convo … damn!
The fascinating this is that the LLM is not acting as a tool here AFAIk, but very much like a colleague.
I have no knowledge of the domain and have only PhD EE level math knowledge, so maybe my bar is too low.
I think the "But this is not intelligence because it is known math" is not a correct argument. It is unknown how the overall higher intelligence of humans works.
What I do notice however is that LLMs are becoming capable of doing an increasing part of the intellectual work I can do, and usually a lot faster.
Just today I presented an agent framework that can take an informal incident statement and propose infrastructure changes to fix it, all evidence backed. This did nothing I could not to, but it did all 5 test cases in 6 - 12 minutes each. I would have found all of the monitoring indications it did, but it would have taken me a day per test case. The LLM also included sass to silly tickets. ("This is not even worth spending monitoring resources on. It's obviously a configuration problem.")
That's how this is reading to me as well. It's just fast at slogging through a certain level of "simple" transformations.
That argument says very little, emergent behavior is a thing in complex systems with billions of parts.
Humans can also be reduced to voltage potentials propagating along of tubes of fat and synapses getting rewired.
There is clearly intelligence there. We have no way to recognise intelligence other than the appearance of intelligence and this very clearly displays that.
It's also quite clearly different to human intelligence in some notable ways, but not in any that preclude describing it as intelligent. At least for normal non-pedantic definitions of the word.
Everyone uses "intelligence" to mean something slightly different, so for this to be a useful claim to make or refute we need to come up with new, intentionally-pedantic, terms (or new domain-specific definitions for vague existing ones).
Yes, trying to communicate (or watching others try to communicate) about these topics is incredibly frustrating because it's pretty much impossible to make any progress without interrogating people's different definitions, but nobody wants to do that because it would mean being pedantic, splitting hairs, etc.
It's not like this is a new problem. Turing had a definition most of a century ago, he wasn't the first and certainly wasn't the last. I don't think we need new terms necessarily, and I doubt we're all going to agree on a definition tomorrow.
I'm no intelligence researcher or philosopher; but, I think LLMs make us confront the (IMO, now clear) distinction between cleverness (intuition), reasoning (rational argument), and consciousness. I suspect that we think of "intelligence" as either of the first two welded to the latter. In that vein, I'd say that consciousness may be just another emotion: happiness, sadness, egoness.
I'd say an entity capable of instructing one of the leading mathematicians of his era is pretty clearly intelligent by any reasonable measure - however it might be arriving at its output.
I think we have wildly different conclusions about what happened here. You see the machine as instructing Terrence Tao, as if it were Plato teaching Socrates about the theory of forms; I see Terrence Tao using the machine to teach himself, like an intelligent student uses a book. In this case, it's just a book that fools us into believing it can think and reason like we do, because it generates language in much the same way we do when we think and reason.
> that's some Harry Potter kind of "writes itself" book.
I'm sure people thought calculators and, indeed, computers themselves were very Harry Potter as well when they first came out. But in the fullness of time the magic and mystique has drained away, and we're left with the understanding that they're just tools.
> at this point, for me, any comment about LLMs that begins with "it's just ..." is hard to take seriously.
Similarly I have a hard time taking seriously the people who make breathless claims of intelligence where there's only a text calculator with weights applied. It's like watching the devout cry "miracle!" at every strangely shaped piece of toast.
There is no intelligence. If anything, this just shows that natural language and mathematics are both fields which are structured in a logically computable way. And if you have a machine that can compute symbolic logic, you can process both natural language and mathematics.
A second corollary is that rational consciousness and thought is less likely to be contained in language than previously thought, because if language is so simple that a machine can process it, it can't contain consciousness.
If natural language was structured in a logically computable way, we'd have had interesting chatbots by the late 80s, basically as soon as a dictionary fit in local RAM, and for the same reason we got compilers.
Da hole raisin y nat-lang be v. hard is dat i kan rite lik dis an it be cool 4 native engrish speekrs 2 unerstand. LLMs are of course fine with this sentence in exactly the way that Zork's engine couldn't be.
The underlying structure of language, which is grammar, is obviously logical. That the symbols used to represent this grammar can be sometimes fuzzy or ambiguous, is no problem for a machine that takes context and probability into account when translating words to the underlying grammar structure.
It's not "obviously logical", it's a pattern which we mimic to avoid mockery.
example For, semi-randomise I word order can this like, Yoda worse than, and be understood.
> is no problem for a machine that takes context and probability into account when translating words to the underlying grammar structure.
We had to invent Transformers to be able to do that with reliability anything close to being worth caring about. Transformers have to learn from examples, not be pre-programmed.
That's just defining intelligence or consciousness as whatever we specifically can't put in a machine. It's not a very useful definition. (And it's one that completely breaks down to nothing if we do manage to fully put these things into machines or somehow prove it's possible.)
I haven't even been convinced it's fundamentally different from biological intelligence. But it's clearly still missing a few ingredients. But we are really close.
> If AI researchers cared about scientific thinking, they would be intensely focused on the brains of bees.
Basically every academic AI researcher in history was doing what you described. The AI industrialists stopped caring 6 years ago once they realized LLMs seem to have been the only thing in 80 years that actually seems to work at any useful level.
There are plenty of pioneering scientists who are either returning to actual AI research (Yann Lecun, Ilya, etc), and plenty who never left (Richard Sutton) who are doing exactly what you are talking about.
> Basically every academic AI researcher in history was doing what you described.
That is not true. Alan Turing did not view things that way, his test would say that a dog has zero intelligence. Neither did any of the MIT Lispers. And neither do Lecun or Sutskever or Sutton! They are all focused on human intelligence. None of them are even slightly concerned about an AI which is intelligent before it learns any language.
> the only thing in 80 years that actually seems to work at any useful level
This isn't true either! Mathematica / Maple / etc are "old-fashioned AI" and they obviously work. The Lisp expert systems were also useful, though less so than an LLM.
> And neither do Lecun or Sutskever or Sutton! They are all focused on human intelligence. None of them are even slightly concerned about an AI which is intelligent before it learns any language.
??? https://www.youtube.com/watch?v=GvibIstOn_E his arguemtn here is clearly built around using some sort of sensory data to build a model of the world like humans (animals) do. also you clearly decline to mention Lecun who has made this point ad-infinitum
> This isn't true either! Mathematica / Maple / etc are "old-fashioned AI" and they obviously work. The Lisp expert systems were also useful, though less so than an LLM.
i personally find it very strange that non-deep learning AI approaches which essentially boiled down to a giant bundle of if statements, or some very simple statistical modeling were called AI in the first place.
What does "predicting the next token" mean? I ask this every time people say "LLMs are just predicting the next token" and it's maddening that nobody can give a straight answer. Predicting it according to what probability distribution? Every process that produces a sequence of actions (including e.g. a human writing) can be modeled by some probability distribution and therefore their actions are indistinguishable from "predicting the next token" emitted by that distribution.
> to predict the next token you first need to model the universe
Exactly. The "most likely next" series of tokens, for example, when given the first half of a correct mathematical proof, is the correct rest of the proof. I have never seen anyone define "most likely next token" in such a way that this isn't true.
It's like saying that thinking cannot generate new knowledge because all thought is just rearranging the information we get from our senses, or memory of previous information from our senses, and we do nothing more than figure out the most likely word to say next in a conversation.
Either humans are not capable of intelligence or computers are capable of becoming intelligent. Neither or both.
i think people say that thinking that only training to produce the next likely word would end up producing some local minimum word that generally fits but doesn't actually lead to intelligent thought.
that feels like a misunderstanding of how the loss function behaves when used within a sequence
I love this koan and what makes it even better is that is based on a true event: Minsky’s real answer was “It has them, you just k ow what they are yet”
>The second stage of the grant termination process began on March 12, 2025, when Justin
Fox and Nate Cavanaugh – identified in the record as members of DOGE’s “Small Agencies
Team” – met with NEH leadership, including McDonald and Wolfson… Prior to joining the Trump Administration, neither Fox nor Cavanaugh had any
experience in government, public grant administration, private grant administration, or reviewing
humanities projects for scholarly merit… In fact, as both were in their twenties, they did not have much experience
in anything at all – certainly not in anything remotely related to the humanities.
This is wild. Trump administration gave a bunch of young adults access to decide about millions and millions of tax money. No guardrails, no checks, no competence.
It's not wild. This was in the news while it was happening. It was all remarked upon and leaked. If you're just now catching on, I'm very sorry to say you're a major part of the problem
This is a consequence of fascism's adoption of AI and its utter contempt for expertise. Why would we need to hire experts when the robot is magical and all you need to do is ask it like you would ask a genie?
Up and down the Trump admin (and beyond) we see this breathless adoption of AI as if it is already some sort of God.
I read this earlier today and was thinking: how many such mathematically gifted individuals exist I. The world at one time? Assuming there are probably 20-30 Tao-caliber people in the US and an adversarial multiplier of 0.1 (only 1 in 10 such kids are nurtured), we reach 300 for this generation, about 1 in a million.
That means in a generation there are ~ 10k such people in the world. Think about connecting them or nurturing them with AI companions.
* A weightlifter was only awarded a laureate if he were able to lift a heavy stone (have no idea what they were lifting, for illustrative purposes only :-)
* Along comes Archimedes who invents what we would call an exoskeleton. Now any regular guy can lift twice as much as last year’s athlete.
* What to do? You can cancel the Olympiads, but they are actually useful as training, motivation, etc So now you have to give the prize on other factors, eg how well he can lift, has he opened a gym in the city, etc
BTW, physics and bio are not exempt, so those researchers better read and try to stay ahead.
reply