math student
minnesota or california
Joined March 2026
- Tweets838
- Following206
- Followers860
- Likes12.9K
whenever i have to cool some liquid (e.g. coffee or tea) with a spoon i remember the everyday physics stackexchange post where they find the best way to do this. very cool physics.stackexchange.com/qu…
donald fagen would force musicians to work to perfection for hours straight & then awkwardly warble over the track in his strangled ivy league hypernasal dork voice and it Kills because the apollonian instrumentals accidentally give perceptible form to unbearable dionysian Truth
wow so a cost-benefit analysis of the large hadron collider estimates 120% benefit/cost, but only if you count things like tourism to the LHC, the amount of time people spend watching news segments about LHC, LHC website visits, time spent discussing LHC@home, and a survey of how 1,400 college students felt about the LHC.
if you don't count those things it's less than 50% lol
I used to joke to people that LLMs would solve RH in cringe millennial redditor prose (i.e. that we would 'solve math' before making LLMs funny or write well)
Now that this is more or less true (NS instead of RH), how many of us would have bet that ChatGPT would solve a millennium problem before writing a single notable tweet or funny joke? It feels like a weird turn. I don't think we've fully assessed how odd it is that to those of us most familiar with AI prose, LLMs haven't *fully* passed the Turing test yet (maybe this is survivorship bias)
I think also that if LLMs could write "duplicitously", i.e. in different styles, AI safety would have surged much earlier because LLMs would have made the internet unusable. But here we are
maybe math has a beautiful and Serious future but it’s hidden behind ripping off the bandaid & dumping $10 billion of compute onto every open conjecture that exists
So, it seems OpenAI is about to release something like 400 AI-generated proofs on a public server.
To the best of my understanding, people are still trying to figure out how the forced Navier–Stokes example actually works. I know of several experts who are studying it, and it is definitely not an easy task.
Until mathematicians digest that example, however, its impact on mathematics as a living subject seems necessarily limited. What questions does it open? What new ideas does it suggest? What does it teach us about fluids, beyond the yes/no information contained in the theorem itself?
Mathematics does not grow simply by accumulating correct statements. Results have to be understood, connected, explained, challenged, reused. Until that happens, they risk remaining lettera morta.
Now imagine adding 400 more AI-generated proofs. For each of them, until some human mathematician picks it up, studies it, and connects it to the existing mathematical culture, isn't it in essentially the same position?
So my question is genuinely: what is the intended mathematical value of releasing hundreds of proofs at once? What happens if there simply aren't enough people willing or able to read them?
To be clear, I think mathematicians should discover mathematics by any means necessary, including scavenging through artificial, alien-looking mathematical material, since there may be extraordinary things to find there.
But humans are not machines, as attention, understanding, taste, and mathematical culture are scarce resources.
What happens if the rate at which mathematics is generated becomes much greater than the rate at which mathematicians can absorb it?
