Hacker Neue

Preferences
sanxiyn
Joined 15,014 karma
Seo Sanghyeon <sanxiyn@gmail.com>.
  1. sanxiyn
    Ruby also used to use Bison, uses its own https://github.com/ruby/lrama these days.
  2. sanxiyn
    Yeah. There are other fully open models like Hugging Face SmolLM but they are not common.
  3. sanxiyn
    With cargo --offline, Rust has better than average support for offline build.
  4. sanxiyn
    An interesting bit of history: for a long time Rust maintained first party support for Windows XP, after other parts of ecosystem generally gave up. This was because Firefox needed it.

    https://github.com/rust-lang/compiler-team/issues/378 (major change proposal to drop Windows XP support) notes this history and links to other relevant pages.

  5. sanxiyn
    If you look at the actual code, it runs ping -c 5. I agree ping without options doesn't terminate.
  6. sanxiyn
    Yes, but in fact compensating for bad questions is a skill, and in my experience it is a skill excelled by Claude and poorly by Gemini.

    In other words, better you are at prompting (eg you write a half page of prompt even for casual uses -- believe or not, such people do exist -- prompt length is in practice a good proxy of prompting skill), more you will like (or at least get better results with) Gemini over Claude.

    This isn't necessarily good for Gemini because being easy to use is actually quite important, but it does mean Gemini is considerably underrated for what it can do.

  7. sanxiyn
    No, but with tail call (using underlying WebAssembly tail call).
  9. sanxiyn
    > "We find that comments by GPT‑5-Codex are less likely to be incorrect or unimportant" -- less unimportant comments in code is definitely an improvement!

    This seems to be a misunderstanding. In the original OpenAI article, comment here is about code review comment, not comment in code.

  10. sanxiyn
    I am not sure how your sun example relates. Language is not whole of reality, but it is clearly part of reality. Memory engram of Coca-Cola is encoded in billions of human brains all over the world, and they are arrangement of atoms.
  11. sanxiyn
    While the broad point that no, OpenAI and Anthropic are not losing money on inference, is almost certainly correct, the model itself is quite incomplete.

    If you actually want to know, I recommend Inference economics of language models from Epoch AI, which is probably the best public model as of 2025-06.

    https://arxiv.org/abs/2506.04645

  12. sanxiyn
    Yes, models give likelihoods you can compare against. No, you can't do that without drafting, because likelihood of token N+2 depends on token N+1. That is, you get P(is, The capital of France) and P(Berlin, The capital of France is), but for the later you need to give "is" as input, you can't do P(Berlin, The Capital of France _).
  13. sanxiyn
    Terence Tao wrote on "blue team" vs "red team" in cybersecurity and how "unreliable" AI is more suited to red team side. I found it very insightful.

    https://www.hackerneue.com/item?id=44711306

  14. sanxiyn
    I understand it is not your focus, but fuzzing still falls short and there is a lot AI can help. For example, when there is checksum, fuzzers typically can't progress and it is "solved" by disabling checks when building for fuzzing. AI can just look at the source code doing the checksum and write the code to fill them in, or use its world knowledge to recognize the function is named sha_256 and import Python hashlib, etc.

    Hint: we are working on this, and it can easily expand coverage in oss-fuzz even if those targets have been fuzzed for a long time with enormous amount of compute.

  15. sanxiyn
    Maybe their ISPs don't do that. There are many ISPs on the Earth.
  16. sanxiyn
    You can look at Debian's reasoning here: https://wiki.debian.org/ArmHardFloatPort. As I understand, the decision was mostly based on hardware.
  17. sanxiyn
    Google quote from https://riscv.org/riscv-news/2024/10/risc-v-announces-ratifi...

    > "Google is delighted to see the ratification of the RVA23 Profile," said Lars Bergstrom, Director of Engineering, Google. "This profile has been the result of a broad industry collaboration, and is now the baseline requirement for the Android RISC-V Application Binary Interface (ABI)."

  18. sanxiyn
    In my experience long context approach flatly doesn't work, so I don't think this is it. The post does mention "tools for keeping notes and preserving important information to be checked later".
  19. sanxiyn
    FYI: Direct Multipixel Imaging and Spectroscopy of an Exoplanet with a Solar Gravity Lens Mission. https://arxiv.org/abs/2002.11871
  20. sanxiyn
    Yes I have.

This user hasn’t submitted anything.