mswphd

6 ms·

1.
by mswphd 5d ago
tanks have very much run over people to kill them. a wheel can very much be used as a weapon.
2.
by mswphd 5d ago
isn't blitzscaling essentially the same thing, except for when American companies do it (not always within their own country, e.g. spotify, netflix, or amazon)
3.
by mswphd 5d ago
the way LLMs write math is not beautiful. it is exactly analogous to the software that LLMs develop is not beautiful. it may achieve impressive end products, but if you like understanding the methods/architecture, looking under the hoo
4.
by mswphd 6d ago
I won't take a side in things, but OpenAI stated the model they used here started training August 28th. Note that "training" here might mean "post-training with RLHF an Astra base model" or something. but training h
5.
by mswphd 7d ago
worth mentioning there's some indication the 240k peak was massively inflated by bot accounts. but by all means there are many fewer bot accounts this year, and it's still ~150k concurrents frequently. so it's still doing ver
6.
by mswphd 7d ago
I guess I don't understand the issue you're raising. If you want to formalize a non-constructive proof, it remains non-constructive, even if you have a computer check the proof vs a human. As a trivial example, in lean you can wor
7.
by mswphd 8d ago
if Stadlmann used a previous OpenAI product, and Astra was trained off of her chat, and had a comparable approach, then it might be comparable.
8.
by mswphd 8d ago
they're using a new model trained since the prompts happened. They are not denying the other group's solution may have been in their model weights, despite it being unreleased.
9.
by mswphd 8d ago
it's very possible they only had to use the massive compute budget because they were trying to plagiarize his work before he published it though, e.g. autonomously do things in ~7 days what he had likely been thinking about for ~1 year
10.
by mswphd 8d ago
you can add law of the excluded middle as an axiom. See midway down this page https://xenaproject.wordpress.com/2017/10/05/more-easy-lean-...
11.
by mswphd 8d ago
this isn't really true anymore. First, a number of the big results are constructions , not counterexamples . For example the existence of a non-sofic group. It was widely believed that non-sofic groups existed (so it wasn't a &q
12.
by mswphd 8d ago
openAI's claimed solution uses a model trained in the last 2 weeks. The prior work would definitely be included in the training set.
13.
by mswphd 8d ago
both wrong. 1. he was working on the same class of problems. He explicitly mentions they were working to extend their techniques to NS (the same techniques that OpenAI may have scooped somehow), and 2. while he was using LLMs to do it, this
14.
by mswphd 12d ago
eh, lattice-based stuff is the first time public-key crypto can use word-size arithmetic, vs full bigint (RSA), or "just" 256+bit arithmetic. it's significantly easier to get right in a side-channel resistant way. this isn&#x
15.
by mswphd 12d ago
and the more recent (post-quantum) lattice-based stuff can get away with ~16 bit arithmetic (it's vectors of ~512-1024 dimension, but the operations are SIMD-friendly)
16.
by mswphd 12d ago
any cryptography can break at any time. Sometimes "sudden" breaks happen. You can't defend against these, so there (perversely) isn't that much of a point worrying about them, besides using schemes many people have thoug
17.
by mswphd 12d ago
2010 is also Citizen's United.
18.
by mswphd 12d ago
I (and I'm sure others) would obviously agree. just a smattering of the obvious cases * for years people have noticed that many members of congress use privileged information to pick stocks. * lobbying post-Citizen's United has a
19.
by mswphd 12d ago
the researchers from the RSA-250 record have publicly claimed that factoring 1024-bit RSA keys is within reach of nation states. Your 1024 bit key is only "fine" because you are a small fry, not because cryptographers think it can
20.
by mswphd 12d ago
faster hardware could also mean gpu/asic/etc.
21.
by mswphd 12d ago
I also doubt this is leveraging a lean4 kernel bug, but I also do not think that a 13m LoC proof that has not been human reviewed closes the book on our understanding of Fermat's Last Theorem, in part because of the decided possibility
22.
by mswphd 12d ago
as mentioned elsewhere, there was a bug in the lean kernel exploited by AI to prove a false statement roughly a month ago https://leodemoura.github.io/blog/2026-8-1-postmortem-for-ke...
23.
by mswphd 12d ago
not really. it's one of the most difficult ones so far for sure, but pales in comparison to something like the classification of finite simple groups. This was initially "completed" in the 80s. You can see the timeline for cl
24.
by mswphd 12d ago
note that this is exactly analogous to an LLM being able to slop code some demo, but not build something more generally useful/maintainable (say something suitable for inclusion in a standard library).
25.
by mswphd 12d ago
it is a general-purpose programming language. for example, it's standard library allows you to do file io, networking, etc.
26.
by mswphd 12d ago
junk theorems aren't the concern, soundness issues in the lean kernel are the concern. Notably, junk theorems are true . Nobody would debate that the junk theorem is true. The main thing people would say is that junk theorems, while b
27.
by mswphd 22d ago
if you do want to do local LLM stuff, mac studio is significantly better than most linux dev boxes.
28.
by mswphd 23d ago
In general you don’t need things that fancy. Instead, you can take 1. Some known set of architectures, with 2. Some known set of (constant time/variable time) operations And then prove things about programs written against those archit
29.
by mswphd 27d ago
someone tried doing this for rust, but the reaction was that it was vibe-coded/low quality. https://github.com/rust-stdx/stdx https://news.ycombinator.com/item?id=48571266 Note that there are baby
30.
by mswphd 1mo ago
misspoke, meant rim brake bike

More ›