> I think Terry Tao is a great litmus test for AI zealotry (both pro- and anti-). Just in this thread, we have people twisting themselves into knots about how he "sold out" or "not doing math the right way" or whatever. To him, AI is a tool, like any other.

That’s an Anti example. What’s a Pro example?

A smart phone was just a tool at first, but over time society has become overly depedent on them. Most of us are now addicted to our smart phones in one way or another, and that has consequences that play out across society as a whole.

AI not only provides potential to cause society to become overly dependent on it, but it's being developed by/pushed for by the same fucking people who caused our societies smartphone addiction.

Once you recognize what we've lost already, it's hard to turn off your brain and just compartmentalize this away as a "just a tool". Nothing that is adopted so widely is "just a tool," and thinking of it in those terms eliminates the ability to analyze the potential downstream effects it will cause.

I do not know about this but, to be honest, he (or his Dpt, or whatever) has the money and connections to try the hidden-behind-closed-doors stuff.

We mere mortals (I am a prof. of Maths at Uni) do not.

"Tao has sold out to the AI grifters to prop up the AI hype bubble" is not a take I expected to see.

I think we can all be a bit grounded and understand reality as we see it -- one of the smartest living mathematicians is using an important invention. Not necessary to believe in any conspiracy theory.

> I struggle to see the beauty in a gargantuan lean proof constructed by 100 different people

Why does it need to be beautiful? Once you proved it it's true and you can use its consequences in math, sciences and engineerings.

Arguments about beauty don't lead anywhere constructive because they are too observer- and context-dependent. Poincaré himself was decrying continuous non-differentiable functions as abominations. The monster group is, well, just like that. What feels intellectually ugly for one generation is natural for the next, and the field moves on

the analogy with experimental physics is a good one - being sure something is true is a good first step to developing an elegant proof of its truth.

the way to interpret the gigantic lean proof is not by inlining each lemma, looking at all the lines, and thinking "yeah that's a lot". That's also not the way to read a paper.

Instead, you proceed in layers of abstraction. For example

1. the main claim may rest on some set of sub-claims, as well as some local (to teh main claim) work to "patch things together"

2. each of those sub-claims themselves may require other sub-claims + local work, etc

These can be collected into a dependency graph. In lean, this is often called a "blueprint". Here is the blueprint for the formalization of the Polynomial Frieman-Rusza conjecture (now a theorem, by Gowers, Green, Manners, and Tao).

https://teorth.github.io/pfr/blueprint/

This layer of abstractions is (roughly) equivalent a different way to format mathematics. You could remove the Lean component (let alone any AI), and create such a dependency graph for a paper. I would argue this is a clearer way to format mathematics (again, ignoring both the formal verification applications of it, as well as AI).

Any mathematics paper intrinsically has a graph such as this underlying it, and tries to make the various linkages in the graph clear via prose. Prose is only so powerful a way to organize things. I'm sure you're familiar with the way early mathematicians would describe various formula (e.g. the quadratic formula) via prose. It is very hard to understand.

Separately from this dependency-graph perspective, you can do things like

1. add formal verification. Now, each component in the dependency graph is verifiable with high confidence (though harder to write and read). This has some benefits and downsides. Harder to write and read is bad. Being able to have high confidence in the veracity of the result is *very* good. It allows larger collaborations in mathematics. Previously, a large collaboration would require all mathematicians to trust eachother to a large extent. This is (practically) difficult.

2. when each component can now be verified to high accuracy, you can now throw AI at it. I won't extoll the virtue of this. There are parts of it that seem interesting, but many "AI for Math" things currently are stil producing unformalized papers (in prose).

Maybe the main thing I'd say is that this type of "graph structure, with each component trusted" is already implicitly what mathematicians do. You write papers that cite other papers etc. Except now, instead of needing to look for status signals to trust papers (or invest personal effort), you can look for another (honestly fairer) signal to trust papers. So there's a sense in which formalization allows for the democratization of mathematics. I do think there's something beautiful about that.

I think what people find beautiful in math is largely something that enables the mathematics (or physics) to be translated to something that they can think about intuitively, and what people can handle in an intuitive way is largely an artifact of what the brain evolved to be able to think about "naturally". But it's quite possible that most things that are true about the universe or math are just ugly and unintuitive, and the pursuit of truth shouldn't necessarily be limited by what people can easily reason about and hold in their heads.

Beautiful explanations are lovely when they exist, but we shouldn't wait for them if we can also find the truth through an ugly method.

> One could argue for pure mathematics that is of no practical utility

wait what is the math with no utility

Ahem. Define "Pre-AI". Automated theorem proving has been an AI task right from the very beginning with Simon and Newell's Logic Theorist, presented at the Dartmouth workshop in 1956.

Logic Theorist soon proved 38 of the first 52 theorems in chapter 2 of the Principia Mathematica. The proof of theorem 2.85 was actually more elegant than the proof produced laboriously by hand by Russell and Whitehead (2026-03-20: What is called here Theorem 2.85 is, in fact, numbered as 2.53 in the page 107 of the 1963 Cambridge University Press edition (https://www.uhu.es/francisco.moreno/gii_mac/docs/Principia_M...) and which appears, under the same 2.53 number, on page 112 of the 1910 CUP Edition, according to the digitalization on wikibooks (https://en.wikisource.org/wiki/Russell_%26_Whitehead%27s_Pri...)). Simon was able to show the new proof to Russell himself who "responded with delight".[17] They attempted to publish the new proof in The Journal of Symbolic Logic, but it was rejected on the grounds that a new proof of an elementary mathematical theorem was not notable, apparently overlooking the fact that one of the authors was a computer program.[18][17]

https://en.wikipedia.org/wiki/Logic_Theorist#History

Maybe some people only understand "AI" to mean "LLMs" but, particularly in maths, LLMs ain't going nowhere without a symbolic solver (or a human mathematician) verifying their output.

Do you mean RenTech? Not sure how you'd know they're heavily invested in anything given the notorious secrecy of the firm. Maybe their public funds have invested in AI, but their most recent 13F shows a 23% tech sector allocation, and their public funds are maybe only half of their total AUM.

Regardless, doubting the legitimacy of Quanta bc it's a Simons Foundation initiative is foolish.

Tao doesn’t seem to have been all that corrupted by the AI money. He’s signatory to the Leiden Declaration after all.

>A programmer

It won't be a programmer doing this work, because they will have gone the way of the dodo.

It'll be workers specific to a certain domain (e.g. engineer, architect, accountant) doing this on top of their usual work.

The software industry will collapse.

Just like me! I like AI because of how smart I am.

There is more to life than productivity.