Terence Tao

I know that this will probably be down-voted to death but I don't like these hyperbolic takes. I know that Sabine did use this title for click-bate purposes that she is now mostly doing YouTube videos (she had horrible experience that unfortunately not rare in scientific community [1]). I understand that the field of particle physics which is the corner stone in fundamental physics is not showing the great advances that it used to have a couple of decades ago. But I think people really don't understand that the field is still advancing and although these advances are less catchy to be reported in mainstream (and don't get traction if posted on HN) it is not dead or dying.

There is a reason why we had a particle data group updating the PDG [2] each two years (you can order physical copies for free but please don't do if you don't need one). People were writing about that since after the big discovery of Higgs boson (that was 12 years ago). We still have a lot of measurement and puzzles that is less about unification theory that people usually would talk about. Theory people are coming up with all different ideas even if some are not testable now but that job of theorist is mainly come up with ideas and help bridge the gap later.

I would suggest everyone interested in this topic to read the electroweak current chapter of the book called "How experiments End" [3] to understand a historical example to how we approached the standard model when it was first proposed. Most of the…

Tao has a comment relevant to that question:

"I doubt that anything resembling genuine "artificial general intelligence" is within reach of current #AI tools. However, I think a weaker, but still quite valuable, type of "artificial general cleverness" is becoming a reality in various ways.

By "general cleverness", I mean the ability to solve broad classes of complex problems via somewhat ad hoc means. These means may be stochastic or the result of brute force computation; they may be ungrounded or fallible; and they may be either uninterpretable, or traceable back to similar tricks found in an AI's training data. So they would not qualify as the result of any true "intelligence". And yet, they can have a non-trivial success rate at achieving an increasingly wide spectrum of tasks, particularly when coupled with stringent verification procedures to filter out incorrect or unpromising approaches, at scales beyond what individual humans could achieve.

This results in the somewhat unintuitive combination of a technology that can be very useful and impressive, while simultaneously being fundamentally unsatisfying and disappointing - somewhat akin to how one's awe at an amazingly clever magic trick can dissipate (or transform to technical respect) once one learns how the trick was performed.

But perhaps this can be resolved by the realization that while cleverness and intelligence are somewhat correlated traits for humans, they are much more decoupled for AI tools (which…

(flailing here, to get Tomte in on this, don't mind me :)

supportive case studies abound in open source-- Valve saving desktop Linux most obviously..

(Possible "the-house-that-Jack-built" counterexample is, ofc, Linus & his kernel-- it even screws with the Microsoft latent framing! Srsly, what's the monystory there?)

More productively: riffing on Jobs' "if you care about sw you must work on hw", here's another "pecuny-agnostic" approach

If you care about turning academics self-sufficient, you must help the self-sufficient pursue academics without disappearing into the usual failure modes (quackery, irrelevance, Bryan Johnson, Sam Altman, PG, Peter Thiel?!?! etc) The organizational issues are much the same, the relative stakes are .. interesting[0]. See the fraught debate following Terry Tao's very recent tacit endorsement of BigTech*

https://mathstodon.xyz/@tao/114586574834318736

*I gather that Tao not keeping those livecoding tubes on tap may indicate that he's seriously considering the issues raised therein

[0] .. than in the case of sports[1]: the mass appeal of μάθημα, to be (insanely?) charitable, is debatable, but the power potentiality, esp. when integrated over time[3], isn't[4].

(Sorry if Grothendieck/Teddy was channeled, not too inadvertently..)

[1] blissfully ignoring the uh not insignificant cases of Olympic horsemen turning into field marshals

[3] The Altman interval, of elite historiographers (or history-makers): exactly one thousand years

[4]…

Great set of observations, and indeed it's worth remembering that the specific details of assistance and setup make a difference of several orders of magnitude. And ha, he edited the last post in the thread to add this comment:

> Related to this, I will not be commenting on any self-reported AI competition performance results for which the methodology was not disclosed in advance of the competition. (3/3)

(This wasn't there when I first read the thread yesterday 18 hours ago; it was edited in 15 hours ago i.e. 3 hours later.)

It's one of the things to admire about Terence Tao: he's always insightful even when he comments about stuff outside mathematics, while always having the mathematician's discipline of not drawing confident conclusions when data is missing.

I was reminded of this because of a recent thread where some HN commenter expected him to make predictions about the future (https://news.ycombinator.com/item?id=44356367). Also reminded of Sherlock Holmes (from A Scandal in Bohemia):

> “This is indeed a mystery,” I remarked. “What do you imagine that it means?”

> “I have no data yet. It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.”

Edit: BTW, seeing some other commentary (here and elsewhere) about these posts is very disappointing — even when Tao explicitly says he's not commenting about any specific claim (like that of OpenAI), many people seem to be eager to…

>I'm not expecting novel scientific theories today. What I am expecting are signs and hints of such genius.

Like I said, what exactly would you be expecting to see with the capabilities that exist today ? It's not a gotcha, it's a genuine question.

>And, I'm sorry, I don't buy the excuse that these tools are not "yet" as capable as the best and brightest humans.

There's nothing to buy or not buy. They simply aren't. They are unable to do a lot of the things these people do. You can't slot an LLM in place of most knowledge workers and expect everything to be fine and dandy. There's no ambiguity on that.

>They contain the sum of human knowledge, far more than any individual human in history.

It's not really the total sum of human knowledge but let's set that aside. Yeah so ? Einstein, Newton, Von Newman. None of these guys were privy to some super secret knowledge their contemporaries weren't so it's obviously not simply a matter of more knowledge.

>Are they not intelligent, capable of thinking and reasoning?

Yeah they are. And so are humans. So were the peers of all those guys. So why are only a few able to see the next step ? It's not just about knowledge, and intelligence lives in degrees/is a gradient.

>If all this is true, surely we should be seeing incredible results produced by this technology. If not by itself, then surely by "amplifying" the work of the best and brightest humans.

Yeah and that exists. Terence Tao has shared a lot of his (and his peers)…

I’m in the skeptical camp. Whatever theory that will eventually emerge will not be as solid as:
1. Theory of pattern recognition (as developed in 80s and 90s)
2. Theory of thermodynamics
3. Theory of gravity
4. Theory of electromagnetism
5. Theory of relativity
Etc. because of two reasons:
1. While half of deep learning is how humans construct the architecture of networks, the more important half relies on data. This data is a hodgepodge of scraped internet data (text and videos), books, user interactions etc., which really has no coherent structure
2. To extract meaningful insights from this much data, it takes models of enormous size like 10B+. The thing about random systems (in the mathematical sense) is that it takes “something” of order of magnitude bigger size to “understand” it, unless there is some concentration of measure type mathematical niceties (as in thermodynamics), which I don’t think is there in these models and data. This is the same reason I don’t think humans will ever be able to “understand” human consciousness. It will take something of an order of magnitude bigger than our own brains to do that.
Here is Terence Tao explaining this concentration stuff in another context: https://mathstodon.xyz/@tao/113873092369347147
I would love to be proven wrong though.