I also feel like many of the current guest posts have overlapping ideas and arguments and so I get less out of each one of them. Maybe a separate blog for endorsed opinion pieces would be appropriate.
The point is that so far the AI is not doing a good job at explaining the "trick", so we (humans) have to do it. And the way these results are published at the moment (that is, dumping a load of proofs with badly written explanations) is not cooperative to enable this distillation (for example, presenting results at conferences and engaging with mathematicians). I recall Tao working through the disproof of the Jacobian conjecture, stating some steps as "miracles" for lack of better terms. If a human solves a problem in an unexpected way, at least there is some reason why they chose this path, which can help in understanding. This is not available to the same extent with AI generated proofs.
> The point is that so far the AI is not doing a good job at explaining the "trick", so we (humans) have to do it.
Sure, but that's a real technical limitation with current AIs, not something that AI firms should be blamed for. And if anything, this creates a viable career path for the mathematicians who were "scooped" wrt. the original solution: they can at least puzzle out what exactly the AI managed to do. Many practitioners are actually quite excited by this possibility; Tao's stance is by no means universally shared.
I'm about to start a PhD in mathematics. The idea of puzzling out what an AI did to prove something sounds quite boring and unappealing. But yes, maybe it's a viable career path.
I'm just not happy with the presentation of OpenAIs result. Maybe they could have gotten into contact with the people of the research areas of the problems that they solved and worked with them to create a better exposition. Sure, it's a slow process and requires lots of staff. But I believe that they can afford it.
> The idea of puzzling out what an AI did to prove something sounds quite boring and unappealing.
The way I see it, it's no different from a lot of grad student work where you have to figure out what a human-written proof is doing.
> Maybe they could have gotten into contact with the people of the research areas of the problems that they solved and worked with them to create a better exposition.
That's what Anthropic is doing, and the issue is that people will complain that they weren't the chosen "person to work with". OpenAI's approach is more like a race where everyone's at the same starting point: they get the AI's raw proof to work on and have to figure out how it works.
... and life moved a bit further in those months. Experiences that could be had, meeting people that could have happened.
If its really relaxing to you, I rest my case. It wouldn't be to me, in same way no computer game or general computer activity ever could. Physical movement, ideally in the nature, the wilder the better. If you live too far reconsider permanently improving QoL by moving closer. Recharges batteries like nothing else.
There is still tomorrow to be more active, right, to catch up. But not with the same mindset that wasted today or yesterday.
I'm a somewhat ambitious climber (mostly indoor, however) and am in general very active. Factorio is an effective way to keep my brain busy as sometimes I get overwhelmed with my thoughts. But of course it is not the only thing that I enjoy and the remark about autumn was an exaggeration. I'm also very much against considering days "wasted". Sometimes I just need to take a day off and spend it at home, as long as there is a balance.
Beautifully put, better than I could have, why I don't touch games in general. They're amazing, but they make you feel like you're advancing, achieving, experiencing, when really you're not, you're barely doing anything than jumping into the future. But, as I understand you to mean as well, not everyone is like that, for some it feels to bring balance. In particular, if you're a coach, surely you have no shortage of human interaction, or being outside, doing sports, and networking. But it's not for me for sure.
I'm not sure if you do yourself a favour with that ambition. If you enjoy games from time to time, you're not "wasting" your life. You are doing something that fills you with joy. Striking a balance is of course important, but allowing yourself to relax from time to time and just have a little fun is too.
For sure. Just like movies. Everything in moderation. But I won't buy 60 bucks AAA games, they're either a time sink, or a waste of money. Indie is enough. I uninstalled Factorio after having spent 90% of my free time on it for a few days.
You're correct that nobody really understands what these huge Lean proofs actually say. However, the initial statement, even for Navier-Stokes, is not very long [0]. Still, you are also right that sometimes the problem statement can be wrong but it is highly unlikely here.
They said that a customer would have paid around 15 million for the required compute. I can't imagine that this was not a significant internal spending even with "free" tokens.
At Astra API prices that's 300B tokens (I saw 130B output tokens claimed elsewhere), large but not unheard of if you consider it across a few people doing random experiments with best-of-n type things. On my personal account, I've done a billion+ token days just on a normal pro 20x subscription. I know many others that wildly outpaced that by orders of magnitude. This was apparently 130B over 89 hours, so about 30x that rate. When things are free and you're expected to token max 30x seems fairly reasonable to me.
If you consider this as a cost to be compared against the question: "What does it take to be able to prove that you have a model that can solve the hardest problems that humans know about?", then spending a some amount of thousands/millions to know the boundaries of that seems not too important in comparison.
You've also got to consider this as compute that's allocated to pushing the frontier of what models can do, so while it's using GPUs that have been paid for etc., it's not like it's a cost that's supposed to be use less of this so that others can have capacity. If you made researchers afraid to use capacity like this, a lot of the things that improve would tend to do so significantly slower. (some may say that's a good thing ;)
A good way to think about this is when tokens are free, you get to choose whether you're optimizing for latency or intelligence rather than having to consider price.
If mathematicians aren't solving problems people are having (which your comment seems to imply), then putting them out of their job with AI is not a bad thing. Of course mathematicians are solving problems, just in a very different way than other professions.
Steam engines actually replaced what a horse does with a machine that could do the same job. So far LLMs aren't doing this - they're producing Lean proofs but very little to actually aid in understanding (again so far pretty much all of these proofs have been extremely difficult optimizations of known techniques). Mathematicians do solve a problem that people have: they build theories that explain the world and give us mental models to navigate questions in science, technology, etc., this just isn't a problem that LLMs solve.
So the issue isn't so much that LLMs will replace mathematicians, but that AI companies bragging constantly about how their machines "solve math" will convince people who don't understand the value of math research to no longer fund it, or students who don't yet understand why learning math is useful for developing their brains that it's a waste of time. That could put mathematicians out of a job without providing a useful replacement.
Motto: a mathematician's job isn't to solve the Hodge conjecture, it's to understand why the Hodge conjecture is or isn't true, and turn that understanding into something that makes it easier for the next person to grasp/use/enjoy.
LLMs absolutely have the potential to make this job easier, but the way in which these companies are using them right now risks being antithetical to that goal.
You can also think of conditional convergence as convergence under the condition of a specific order. It then turns out that unconditional convergence (that is, convergence where it doesn't matter what order you choose) is equivalent to absolute convergence (that is, the sum over the absolute values converges).
https://news.ysimulator.run/item/2297