cs702

11 ms·

211.
by cs702 1y ago
Not necessarily. It depends on the second-order economic effects of these investments in AI. For example, companies that build new giant data centers for AI pay salaries to thousands of temporary construction workers who then spend some o
212.
by cs702 1y ago
Key quote: > The hundreds of billions of dollars companies are investing in AI now account for an astonishing 40 per cent share of US GDP growth this year. In other words, investment in AI is a massive private-sector stimulus program
213.
by cs702 1y ago
I disagree. My argument could not apply to other companies, because AMD is the only other maker of GPU hardware that could conceivably compete with Nvidia at scale in the AI market. Intel's GPUs have never been in the same league. Ne
214.
by cs702 1y ago
OpenAI and AMD both have strong incentives to chip away at Nvidia's dominant position, and the best way to do that is by releasing software tools to the public. Ask yourself: What would first-class PyTorch/Triton and Jax support f
215.
by cs702 1y ago
Today, yes, it's true that AMD hardware can be competitive mainly for inference. The story here is about the future . Over time, OpenAI would benefit if AMD hardware becomes competitive for training too. Nvidia currently gets to charg
216.
by cs702 1y ago
OpenAI now has an option to buy 10% of AMD for $0.01 per share, subject to certain milestones. Basically, AMD is giving up equity to buy its way into the AI market. It's fantastic news , because OpenAI and AMD will now work together t
217.
Advice from nixCraft(mastodon.social)
1 points by cs702 1y ago|0 comments
218.
by cs702 1y ago
Brilliant. This deserves its own HN submission. I submitted it but it was flagged due to the title. Thank you for sharing it on HN.
219.
by cs702 1y ago
Hmm... I'd say most AI emperors have no clothes -- that is, no profits and no prospect of them in the foreseeable future.[a] A tiny number of them are actually wearing something, like a fig leaf. They are the ones won't be cau
220.
by cs702 1y ago
Correction: 4.3B in revenues . Other than Nvidia and the cloud providers (AWS, Azure, GCP, Oracle, etc.), no one is earning a profit with AI, so far. Nvidia and the cloud providers will do well only if capital spending on AI, per year, rem
221.
by cs702 1y ago
The Economist's analysis concludes that the bulk of exorbitant rents in the US is captured by providers of health-care services who take advantage of the healthcare system's opacity: * hospitals, * insurers, * pharmacy-benefit man
222.
by cs702 1y ago
The OP was submitted before, by the author, but it failed to catch on: https://news.ycombinator.com/item?id=45169959 I think it will be interesting to HN, so I'm submitting it again.
223.
Benjie's Humanoid Olympic Games (A Competition for Humanoid Robots)(generalrobots.substack.com)
1 points by cs702 1y ago|1 comments
224.
by cs702 1y ago
On one hand, it's just Linus Torvalds calling it like he sees it, without giving too much considerations to things he deems less important, like, say, how the recipients of his message might feel. On the other hand, as always, he makes
225.
by cs702 1y ago
See also: https://x.com/jasonfurman/status/1971995367202775284
226.
Does GDP growth minus AI capex equal zero?(ft.com)
2 points by cs702 1y ago|1 comments
227.
by cs702 1y ago
That's for a small and shallow neural network. I was wondering about LLMs specifically.
228.
by cs702 1y ago
Well, engineers building physical systems like airplanes and rockets use Lyapunov exponents to avoid chaotic behavior. No one sane wants airplanes or rockets that exhibit chaotic aerodynamics! Has progress stalled in this area? I don&#x27
229.
by cs702 1y ago
Interesting. Nothing prohibits us from thinking of pretrained LLMs as dynamical systems that take a token state and compute an updated token state: x_{n+1} = LLM(x_n) , starting from an initial token state x_0 . Surely we can compute traj
230.
by cs702 1y ago
Yeah, lately I've found myself editing out my em-dashes , because otherwise people may think it's an LLM writing, not me. Sigh.
231.
by cs702 1y ago
Basically, it's the "rapture" as a Marvel-style fantasy-action movie, generated by AI. Video: https://www.youtube.com/watch?v=f3uhVJTxNko Original source: https://www.npr.org/2025/09&#x2F
232.
by cs702 1y ago
Also, from that page: https://towardsdatascience.com/attractors-in-neural-network-...
233.
by cs702 1y ago
@dang, thank you for doing that. When I posted the OP, I considered changing the title, but decided not to editorialize it, per the guidelines.
234.
by cs702 1y ago
Cool stuff. I've been meaning to give this a whirl since I saw it here. I just tested the code for estimating Lyapunov exponents in parallel and it worked on the first try. Nice! These "goom" thingamajigs could be useful for
235.
by cs702 1y ago
The OP shows examples of people being unable to solve a problem in Julia that they solve quickly after switching to PyTorch, Jax, or TensorFlow, so the OP is implicitly recommending those alternatives.
236.
Random Attractors – Found using Lyapunov Exponents (2001)(paulbourke.net)
137 points by cs702 1y ago|28 comments
237.
Correctness and composability bugs in the Julia ecosystem (2022)(yuri.is)
105 points by cs702 1y ago|79 comments
238.
by cs702 1y ago
Also beautiful. Thank you for sharing it on HN.
239.
by cs702 1y ago
Beautiful. I always find it a bit shocking that there is so much structure in the states of chaotic systems as they evolve over time. Thank you for sharing this on HN.
240.
by cs702 1y ago
Bubble or not, AI is eating everything , not just YC. In particular, I've noticed the topic of AI is eating up corporate meetings, turning them into a kind of human slop.

More ›