Live data from Hacker News

Viewing profile — anematode

anematode

HN member
Joined
Wed, May 22, 2024, 8:01 PM UTC
HN karma
758
Public activity
246 items

About anematode

Opinions expressed here are solely my own. Website: https://anemato.de

Recent public activity

  1. comment
    Comment #49199055

    Yes, that's one approach. Another reasonable approach is to get out a mask from the comparison using `vmovmskpd` and use that to look up a shuffle constant, since there are only 16…

  2. comment
    Comment #49194069

    Great link, nice find!

  3. comment
    Comment #49193833

    Good question. I personally doubt that the compress instruction is easy to coax compilers into generating, as there are many edge cases to consider. For example, you'll notice here…

  4. comment
    Comment #49192285

    Nice post! You can do even a bit better if you're willing to use intrinsics. In particular this kind of operation is well-suited for compress-type operations, available as a first-…

  5. comment
    Comment #49192097

    That's awesome :) Do you have any links to your guys' work that I could check out?

  6. comment
    Comment #49188684

    It's a dog-eat-dog world, eh? :) Anyway, I don't see how this relates to the original discussion. Nor do I think the irony is particularly deep: As far as I'm aware, Deep Blue didn…

  7. comment
    Comment #49188572

    :EmotiTrumpet:

  8. comment
    Comment #49188433

    > Some times that's appropriate, like if there's a competition for hand-made or human-written something. Indeed, and this is the problem: unlike woodworking, where my power-assiste…

  9. comment
    Comment #49188366

    My implicit request, which I should have made clearer, was for him to delete (or make private) the repository and move on to greater ideals.

  10. comment
    Comment #49188315

    You'll notice I didn't mention copyright, and that's because I don't think it's relevant. I think your analysis is correct. Keep in mind too that Stockfish – the engine I work on, …

  11. comment
    Comment #49187647

    I'm the OP of the cited thread, ask me anything :P

  12. comment
    Comment #49177851

    Meanwhile: https://textslashplain.com/2026/08/04/security-is-hard-yall/ > The Cloudflare folks apparently want security issues reported via HackerOne (which wouldn’t let me log in …

  13. comment
    Comment #49132310

    Fully agreed. As someone who both loves chess and works on chess engines... these comparisons to chess needs to stop.

  14. comment
    Comment #49064986

    > deferring int optimizations for later This part made me laugh out loud

  15. comment
    Comment #49052035

    https://en.wikipedia.org/wiki/Incitement#United_States If done obliquely enough... e.g., just making a tracker and having the ability for people to report broken cameras, you're pr…

  16. comment
    Comment #49050758

    A raw and powerful piece, thanks for sharing. For years I've wanted to get back into self-studying mathematics, not to make serious contributions but just to appreciate its beauty;…

  17. comment
    Comment #49025650

    For security?? Why else would they purposely lower cyber refusals in that case.

  18. comment
    Comment #49002116

    > Do you think there is such a thing as perfect security? No one can "get it right" in the face of arbitrarily high intelligence Why didn't they run the model against the sandbox f…

  19. comment
    Comment #49000410

    Disgusting.

  20. comment
    Comment #48971332

    Same observation. One thing you can do is close out the session and `claude --resume`

  21. comment
    Comment #48971321

    It's a "democracy", in which some people (Anthropic and co., and anyone willing to suck up to them) are more equal than others...

  22. comment
  23. comment
    Comment #48909829

    Demis sold out. "Harnessing AI safely" means nothing when your technology is used to help the US government kill and surveil people.

  24. comment
    Comment #48889377

    Brilliant! Hadn't seen this technique before.

  25. comment
    Comment #48886955

    Sure, but you'd use a low-precision approximation for that; you don't need full double precision