nilkn

5 ms·

31.
by nilkn 2mo ago
It's well understood at this point that China's results are simply not purely the result of distillation. They're the result of significant investment with a willingness not to make a profit any time soon. That's why dis
32.
by nilkn 2mo ago
You can disagree, but simply stating disagreement isn't substantive. Linux can be enormously important while also not being subject to exponential value creation through improvements to itself.
33.
by nilkn 3mo ago
Linux is not a relevant comparison, because the fundamental capabilities of an OS do not grow exponentially with time or investment. If building full sovereign ownership of the production chain for frontier-level AI were so easy, there woul
34.
by nilkn 3mo ago
The keyword in your sentence is "access". The rest of the world would have cheap access to AI, but no ownership of anything at the frontier. If the US joined the rest of the world, declined to develop frontier AI, and allowed it
35.
by nilkn 3mo ago
You're not seeing it because you're thinking in terms of internal US economics, but this is a question about international economics and geopolitical leverage in the distant future. The reason to consider the ban is because it mig
36.
by nilkn 3mo ago
Perfect -- that's great to see. The proof strategy in Lean appears essentially identical to the natural language strategy (as much as is reasonably possible). I think this settles it!
37.
by nilkn 3mo ago
We're talking past each other for some reason. I'm not "unhappy" with anything. I just pointed out that (1) a result like this requires peer review by a professional human mathematician, which fundamentally bottlenecks p
38.
by nilkn 3mo ago
You are assuming that the latter, once autonomously discovered and verified at scale, could not simply be translated into the former, also perhaps autonomously at scale (or otherwise selectively as determined by human interest, taste, and r
39.
by nilkn 3mo ago
I disagree. It's the only way to scale AI mathematics far beyond human mathematics. Any interesting verified result would, obviously, be rewritten back into natural language for human understanding and consumption (as well as potential
40.
by nilkn 3mo ago
Nah, if it produced the proof in Lean which is automatically verified to be correct, you could then just write a natural language version of the proof to accompany it (often using AI to do that part too). That's becoming the standard f
41.
by nilkn 3mo ago
Since this isn't in Lean and it's extremely easy for something like this to contain a subtle mistake, I think I'd prefer this be announced by a professional mathematician. The proof appears relatively short and elementary (no
42.
by nilkn 3mo ago
I'm not sure how meaningful this is. Fable only just recently become more broadly available, and GPT-5.6 is launching broadly today.
43.
by nilkn 3mo ago
Codex has arguably been better than Claude Code for months now, but it's flown under the radar because it just didn't capture the same viral marketing effect and OpenAI in general has had more optics / PR issues than Anthropi
44.
by nilkn 3mo ago
There's a lot of really important software out there where being able to easily verify effect-free core logic would certainly be very useful. An e-commerce web app is not a good example. Anything safety-critical -- aerospace, defense,
45.
by nilkn 3mo ago
So the frontier will just decisively shift to open Chinese models in the near future, and once that happens, there will be no catching up.
46.
by nilkn 4mo ago
That's because the company likely doesn't view it as a mistake. The executives did their job: they tried something the company likely considered reasonable (or even strategically necessary) and pivoted based on results. At the exe
47.
by nilkn 4mo ago
I use Windows on my home PC simply because it's such a versatile setup thanks to native GPU support, game support, and WSL2. It feels like I can do almost everything on one machine with little to no compromise. If I didn't care ab
48.
by nilkn 4mo ago
I'm not sure why we would assume that AI-generated or AI-assisted mathematics would never amount to anything useful in the real world. I would expect the opposite: the usefulness and explanatory power of mathematics has been riding an
49.
by nilkn 4mo ago
An inscrutable 1000-page Lean proof may have low transmissibility amongst humans, yet extremely high transmissibility amongst AI mathematicians. Probably AI mathematics needs a specially constructed or trained translation or compression sys
50.
by nilkn 4mo ago
I don't think Terence Tao sold out. However, just looking at it from OpenAI's perspective, this kind of advertising is almost certainly worth at least one order of magnitude more than $3M to the company.
51.
by nilkn 4mo ago
I bought it on launch day, and I still use it at least a couple times a week. I also pretty much always take it with me when I travel (along with my MBP). Frankly, I'd use it even more, but because it's a fairly anti-social device
52.
by nilkn 4mo ago
Their point in this case is that a wavefunction is spread out over space, which would cause it to be subject to a local clock gradient in curved spacetime. If you wanted to use particles, you'd need to use a Feynman-style "integra
53.
by nilkn 4mo ago
The original point of the stock market was to fund gigantic society-level projects (like railroads). Modern VC has replaced some of that at smaller scales but not all of it at the largest scales. So this could just be the stock market perfo
54.
by nilkn 4mo ago
The beauty of a useful analogy is it allows you to make correct inferences, even without a full understanding. If you view gravity as a mattress, you're stuck. There's nothing to do with it that you couldn't already do, becau
55.
by nilkn 4mo ago
That is a harder question than it sounds. The answer might actually be "there are no such examples", but I'm not confident enough to jump to that with any certainty by any means. Near and inside black holes, there can certain
56.
by nilkn 4mo ago
That's right -- the atmosphere stays attached to the Earth mostly thanks to gravity, and the Earth's gravity in GR is almost entirely the gradient in clock rate near the Earth. Near Earth’s surface, clocks lower down tick very sli
57.
by nilkn 4mo ago
It means clocks tick at different rates depending on where they are. Imagine spacetime as a field of local clocks. Far from the Sun, clocks tick faster. Near the Sun, clocks tick slower. A freely moving object tries to follow the straightes
58.
by nilkn 4mo ago
"Einstein cast gravity not as a force but as the geometric bending of space and time. In a popular analogy, the fabric of space-time is like the flat expanse of a mattress, and a massive object like a star is like a bowling ball sittin
59.
by nilkn 4mo ago
I've actually thought about this. My personal interpretation (which I admit is absolutely a bit playful, but I also find it very fulfilling) is that gravity acts like a resource allocator. It clumps matter together into stars, planets,
60.
by nilkn 4mo ago
That is not unexpected. There are three forms worth considering, actually: bacteria, archaea, and eukaryotes. The first two share DNA/RNA replication, but they operate in completely different ways. The third is dramatically more comple

More ›