I think his point is that AI is not creating new problems. It may solve "the Riemann hypothesis" but may completely fail to posit a "Mythos hypothesis" which is vital to advance the field. In fact, achieving the former may make the latter even harder because it will disincentivize production of human mathematics which has till now been the only source of "interesting" problems.
FWIW this is my understanding of his argument and I am not a mathematician.
As Tao points out, merely suggesting new open questions isn't really sufficient. Part of what gives these problems their fame is their notoriety, their difficulty, the fact that many prodigious mathematicians have spent an evening or week or month or several years studying it.
It wouldn't be as interesting if it had just been solved by the fifth random mathematician who considered it
Notably, gardening a new field of study in math is somewhat nontrivial. You have to introduce the field, illustrate some relevance or connections, and then - and this is key - not solve all of the low-hanging fruit yourself! Because you need somebody else to become an expert in that particular field.
The analog in programming is: if a large company merely open sources a product that's decent but not great and in a language nobody wants to maintain, but they don't commit to maintaining it themselves.
Suddenly there's a bit of a vacuum because in order to provide something of value, you either need to:
1. Implement something more complete than was initially open sourced
2. Or maintain something in a horrendous language while incrementally improving it and keeping it relevant
3. Or rewrite it into a tolerable and maintainable modern language.
What the large company has done is create a vacuum in the tool space where you now require extreme motivation to get someone else to step in.
Note that in this scenario, in 2026, it's actually not such a big deal. I think several recent models could happily translate it into a more maintainable language themselves or happily maintain it in the original crufty one. And so the question is: which parts of this analogy are true in math, too?
Yes, these things are definitely being conflated and there's a fourth conflated thing I think Tao is particular getting at:
(4) Mathematics, the community which is a living system that decides what is interesting, constructs shared frameworks, transfers ideas between domains, develops taste, teaches new mathematicians, and continually emphasizes what counts as important mathematics, especially in it's overall value to humanity.
This is the level in which theory building, simplification, integration and applications are built on. Contributing meaningfully here requires much more than generating solutions - it requires direction and restraint.
Tao seems to be saying that this direction and restraint is the scarce resource that drives mathematical progress, not problem-solving ability. And AI is not only insufficient at it, but results generated by AI are destroying human ability to excercise this resource.
This seems similar to another problem AI sucks at - drafting legal agreements. Despite being great at evaluating, interpreting, and comparing legal agreements it fails amazingly at drafting them. This is because what you don't say/do is vastly more important than what you do say/do.
And AI is great saying and doing things, it's the selection based on implied values that it struggles with.
You pick your axioms and you see what must be true. With math and logic, we can reason about any possible universe even though we live in only one of them.
The sphere of human comprehensible mathematics is finite. Once everything is solve it is not necessary to advance the field. The recurring error her is to say ai is not the product of human effort but another agent. Ai is human. Ai may well be speeding up human comprehension of math to its limits in which case there is no further need to advance the field and mathematicians might need to get a job. Why is this a bad thing?
Most mom and pop stores (Kirana stores) in India cannot get a PoS device easily. UPI solves this problem by being much easier for merchants to onboard and start accepting payments with just a QR code instead of a PoS device.
Not just about PoS device. During early credit/debit card days, lot of stores would pass on merchandise fees directly to customers, giving discount in case of cash transactions, some stores still charge extra. This is not allowed in west as credit cards require same price as cash transactions, and most users perceive credit card rewards as something positive as one would lose out on this rewards when using cash. In India perception is that one would be making loss* if using credit cards. UPI with no fees as of now, is perceived as equal to cash by most people not dealing with black money.
It’s also available via public libraries in USA via Libby if your local library system pays for a subscription, so it’s a way to support the magazine indirectly, since your local taxes pays for your library. The downside for weekly is you have to read it that week, no archive access.
The New Yorker posts digital articles in advance of the release of the print edition.
At the bottom of this article it says: Published in the print edition of the April 13, 2026, issue, with the headline “Moment of Truth.”
As someone who reads the print magazine every week, I always scroll down to check if the article will be published and skip it if so (so I can read it when my magazine arrives).
Thanks. Can confirm the print edition is assigned a date that is 5 days after it becomes available in the Press Reader app in the UK (not sure why) and the article is in that edition.
This looks interesting. What I don't understand is: how was it implemented without a server relay. I am no expert in WebRTC (or P2P for that matter) but I always assumed that there needs to be a central location for users to exchange their addresses and only then a P2P connection can be established. This must be the case here as well right? Or am I mistaken?
That is exactly the case, there is a server in the background responsible for maintaining sessions and setting up WebRTC connections(ICE handshakes), when all clients are connected - messages are sent p2p
> This project was born out of the need for a lightweight analytics tool to track internal services on a resource-constrained VPS. Most SaaS analytics products either lack the scalability or exceed their free tier limits when tracking millions of events per month. Minimalytics addresses this gap by offering a minimalist, high-performance solution for resource-constrained environments.
I recently did a Show HN which you can find on my profile.
FWIW this is my understanding of his argument and I am not a mathematician.
reply