Hacker News: atty

New comment by atty in "Discovery Loop"

atty — Thu, 06 Aug 2026 00:32:30 +0000

Work for a manufacturing company, and we spend a lot of time creating surrogate models for physical simulations. Huge speed ups, 100-1000x possible. You end up still using the “real” simulation for validation, but you can run orders of magnitude more simulations for early design and refinement first.

New comment by atty in "Android may soon restrict on-device ADB"

atty — Sat, 25 Jul 2026 12:59:57 +0000

Sure, sometimes people get social engineered into taking money from their bank account and giving it to criminals. Should banks stop allowing withdrawals?

New comment by atty in "Professor denounces mass AI fraud on an exam at Brown"

atty — Mon, 29 Jun 2026 00:25:27 +0000

And physics

New comment by atty in "Political bias in AI: Where the AI models stand"

atty — Thu, 25 Jun 2026 14:50:16 +0000

Factuality is orthogonal to political leaning generally. People can use the same set of facts and come to very different conclusions. That’s a separate issue from “are these facts correct” and what happens when an individual or entire party starts getting most of their news from highly partisan and unreliable sources.

New comment by atty in "US holds off blacklisting DeepSeek, more than 100 firms deemed security risks"

atty — Thu, 18 Jun 2026 01:51:27 +0000

In most cases it seems (at least to me and colleagues) to be turning out that picking best intelligence is a better option than picking better intelligence / dollar, assuming you can afford the cost. At least on interesting problems. If you’re doing generic web dev work, probably not the case.

New comment by atty in "ICE Awards $25M Iris-Scanning Contract to Bi2 Technologies"

atty — Sun, 24 May 2026 02:06:05 +0000

The federal government and the executive branch has a well specified set of responsibilities to the people of this country, and they are massively failing at just about every metric. So yes, they are truly incompetent.

New comment by atty in "Electrobun 2.0 will be decoupled from Bun due to the Rust rewrite"

atty — Sun, 24 May 2026 02:02:26 +0000

> as long as the tests pass

To be pedantic, tests prove that the code passes the test suite, nothing else. They do not prove by themselves that the code is correct, secure, maintainable, efficient, etc. Those are much harder to measure and have a ton to do with organization, architecture, culture, shared knowledge of the maintainers, etc. All of which is lacking during and after this rewrite.

New comment by atty in "Cloudflare outage on February 20, 2026"

atty — Sat, 21 Feb 2026 19:29:03 +0000

I do not work in the space at all, but it seems like Cloudflare has been having more network disruptions lately than they used to. To anyone who deals with this sort of thing, is that just recency bias?

New comment by atty in "Claude is good at assembling blocks, but still falls apart at creating them"

atty — Fri, 16 Jan 2026 00:36:02 +0000

I think both your statement and their statement are too strong. There is no reason to think LLMs can do everything a human can do, which seems to be your implication. On the other hand, the technology is still improving, so maybe it’ll get there.

New comment by atty in "Someone at YouTube Needs Glasses: The Prophecy Has Been Fulfilled"

atty — Wed, 26 Nov 2025 05:46:42 +0000

Never ascribe to malice what can be sufficiently explained by incompetence. And i think it’s fair to say the best and brightest at Google aren’t turning their attention to YouTube lately. Except maybe to make training datasets for Gemini N+1 :)

New comment by atty in "Uv is the best thing to happen to the Python ecosystem in a decade"

atty — Thu, 30 Oct 2025 02:41:32 +0000

The unspoken assertion that Rust and Python are interchangeable is pretty wild and needs significant defense, I think. I know a lot of scientists who would see their first borrow checker error and immediately move back to Python/C++/Matlab/Fortran/Julia and never consider rust again.

New comment by atty in "Apple Silicon GPU Support in Mojo"

atty — Sun, 21 Sep 2025 23:09:00 +0000

To be fair, triton is in active use, and this should be even more ergonomic for Python users than triton. I dont think it’s a sure thing, but I wouldn’t say it has zero chance either.

New comment by atty in "Chat Control Must Be Stopped"

atty — Tue, 09 Sep 2025 00:13:09 +0000

Which ICE are you referring to? This is an EU law.

New comment by atty in "Vijaye Raji to become CTO of Applications with acquisition of Statsig"

atty — Tue, 02 Sep 2025 21:42:15 +0000

It would be quite strange indeed for Mira Murati to have a say over their technical decisions, considering she does not work for OpenAI :)

New comment by atty in "Ask HN: How can ChatGPT serve 700M users when I can't run one GPT-4 locally?"

atty — Fri, 08 Aug 2025 20:58:39 +0000

The idea that a GPU is dedicated to a single inference task is just generally incorrect. Inputs are batched, and it’s not a single GPU handling a single request, it’s a handful of GPUs in various parallelism schemes processing a batch of requests at once. There’s a latency vs throughput trade off that operators make. The larger that batch size the greater the latency, but it improves overall cluster throughput.

New comment by atty in "Multiplatform Matrix Multiplication Kernels"

atty — Sat, 19 Jul 2025 02:18:18 +0000

I think they’re referring to hardware like TPUs and other ASICs. Which also exist, of course :)

New comment by atty in "How to solve computational science problems with AI: PINNs"

atty — Thu, 23 Jan 2025 01:12:51 +0000

I work on a team that has actually deployed NN based surrogate models into production in industry. We don’t use PINNs for the simple reason that many industrial scale solvers are solving significantly more complex systems than a single global PDE (at least in CFD, perhaps other areas are simpler). For instance, close to the boundaries, the solver our engineers use uses an approximation that does not satisfy conservation of mass and momentum. So when we try to use physical constraints, our accuracy goes down. Even in the cases where we could technically use PINNs we find they are underwhelming, and spending time on crafting better training data sets has always been a better option for us.

New comment by atty in "Ruff: Python linter and code formatter written in Rust"

atty — Tue, 21 Jan 2025 01:45:20 +0000

I do the same thing, and I keep doing it regardless of how many times I see it working properly.

New comment by atty in "TSMC's Arizona Plant to Start Making Advanced Chips"

atty — Sat, 28 Dec 2024 05:46:20 +0000

That is what Intel 18A is, no? In some ways it’s worse than N2, and in some ways it’s better. Overall seems comparable to me, and apparently it’s still on track for next year.

New comment by atty in "How I ship projects at big tech companies"

atty — Tue, 12 Nov 2024 01:25:24 +0000

Generally the game is created by external forces, not the players. If you’re an independently wealthy individual you can afford to work for pleasure. The rest of us? We like to eat and live in comfortable dwellings and take care of our families, so we play the game of pleasing our superiors in return for the money to live a comfortable existence.