Live data from Hacker News

Viewing profile — casparvitch

casparvitch

HN member
Joined
Sun, Oct 17, 2021, 9:44 PM UTC
HN karma
105
Public activity
64 items

About casparvitch

No profile information was provided.

Recent public activity

  1. comment
    Comment #49066384

    Have never seen 27-7, as someone in rest-of-world

  2. comment
    Comment #47527851

    I have been thinking about the difference between 'consumption' and 'creation' style hobbies lately. Spending time drinking different coffee beans, or collecting sneakers, I would …

  3. story
  4. comment
    Comment #46613048

    I agree with this wholeheartedly! Even something a small as fixing up some horrific diverged orphaned git history can help me on a hobby project. And the fact that I see the comman…

  5. comment
    Comment #46611697

    Would love this as an audio visualizer

  6. comment
    Comment #46608721

    I've found in (EU) academia at least that people essentially lie about how much work they do. In anglosphere it's far more common for people to be open/expectant of 80 hour weeks e…

  7. comment
    Comment #46572337

    > A possibly complementary take is that maybe mediums also just... have shelf lives? Like, if there's been no innovation in radio dramas or still life paintings of flowers or super…

  8. comment
  9. story
    AI's Memorization Crisis

    https://archive.md/xitDT

  10. comment
    Comment #46551906

    More inference = more cost (to Anthropic)

  11. comment
    Comment #46550713

    It appears [0] there is now a fix/workaround (?) [0] https://github.com/anomalyco/opencode-anthropic-auth/pull/11

  12. comment
    Comment #46550688

    I doubt you can opt out

  13. comment
    Comment #46550644

    Oh you mean literally claude code, not some hack like opencode? I guess that makes sense yeah, given the tmux interface.

  14. comment
    Comment #46550637

    There's just not that much IP in a UI like that. Every day we get articles on here that you can make an agent in 200 LOCs, Yegge's gas town in 2 weeks, etc. Training the model is t…

  15. comment
    Comment #46550434

    I guess one issue is that you pay $200/month whether you use it or not. Potentially this could be better for Anthropic. What was not necessarily foreseeable (ok maybe it was) back …

  16. comment
    Comment #46550276

    Is it that it's the best coding tool or the best model? I still get the best (most accurate) results out of anthropic models (but not out of CC).

  17. comment
    Comment #46550189

    Yeah fair, I could've chosen better words. My conclusion is: "I don't want to pay for that", not "they shouldn't be able to do that".

  18. comment
    Comment #46550159

    Unsure of the other competition, but I can vouch for synthetic.new's subscription for GLM (+ other open models). Note quite as accurate as Anthropic's models but good enough for ba…

  19. comment
    Comment #46550123

    IDK if Anthropic wants to offer a service at below cost, I don't think they should gate keep which client you access that service over. Or in other terms, I won't use a service tha…

  20. comment
    Comment #46550093

    Maybe Anthropic saw Yegge's gas town concept and got scared off?

  21. comment
    Comment #46296328

    > My very personal opinion is bureaucracy, fat bureaucracy, damned bureaucracy, expensive bureaucracy: believing you can have a better world by regulations IS stupid, and so on... …

  22. comment
    Comment #46296048

    > There's an entire parallel scientific corpus most western researches never see. Most of the scientific corpus (I would wager >99%) is not read by anyone. At least in physics, all…

  23. comment
    Comment #46269051

    recutils is aimed at database rather than config use-cases

  24. story
  25. comment
    Comment #46201335

    Statistical algorithms always make more concrete assumptions of the signal. DFT / Fourier transforms are great as they are a direct mathematical operation, that maps neatly to (bas…