Live data from Hacker News

Viewing profile — fwipsy

fwipsy

HN member
Joined
Sun, Jan 26, 2020, 4:49 AM UTC
HN karma
1,370
Public activity
559 items

About fwipsy

No profile information was provided.

Recent public activity

  1. comment
    Comment #49259460

    To me it indicates that he knew how easy it is to lose your shirt, likely from the examples of his peers -- and then he did anyways.

  2. comment
    Comment #49259408

    > In recent times, Johannes Heidecke, Head of OpenAI's Safety Systems team, also left the company – as did OpenAI’s Chief Futurist Joshua Achiam. It's OK though, they don't need an…

  3. comment
    Comment #49259135

    Why do you think cloud AI doesn't also use these optimizations? They are also compute-constrained. I expect optimization and scale have a multiplicative effect. Surely all the seri…

  4. comment
    Comment #49252969

    The robot market would have to be pretty damn valuable, to be worth throwing away a lead in the model market. Say, 10-100x as valuable.

  5. comment
    Comment #49252762

    It's pretty ironic to read about Twain losing his fortune to bad investments, after satirizing sketchy investments in "The Gilded Age."

  6. comment
    Comment #49251813

    14mb? More like sub-$20 devices.

  7. comment
    Comment #49244378

    Maybe you're right for smaller models, but for frontier models this is the tail wagging the dog. If China has exclusive frontier model capability in the future, that's an enormous …

  8. comment
    Comment #49243858

    Good points, I personally believe that if/when China takes the lead, they will immediately stop releasing model weights. It only makes sense as a strategy to counterbalance (curren…

  9. comment
    Comment #49235793

    Part of negotiating is understanding when no compromise is possible. Cancelling the rental seems like the best outcome here, although I don't like the author's tone much.

  10. comment
  11. comment
  12. comment
    Comment #49235205

    FWIW I've also had success getting AI to write readable code. I guess my point is that allowing it to write opaque code may leave room for unintended behavior.

  13. comment
    Comment #49228500

    I took it for granted that as AIs got smarter, their taste in code would get better. The alternative seems much worse. If AIs get smarter but the code remains opaque, then eventual…

  14. comment
    Comment #49227625

    "We're the good guys because we'll destroy the world a little less."

  15. comment
    Comment #49227613

    I think Anthropic employees think that their product is more like opiates -- highly dangerous, but with a large potential benefit when applied correctly. I don't know what OpenAI e…

  16. comment
    Comment #49227595

    The culture at frontier labs is set by people who have been in the field for over a decade--AI's true believers, who expect it to be a technology as dangerous and disruptive as nuc…

  17. comment
    Comment #49226498

    Well ackshually... > Beaver food flavoring doesn't actually come from the animal's anal gland but from another organ next to it, called the castor sac. I still wouldn't try it, mys…

  18. comment
    Comment #49201293

    CYA. This approach not working is "the user's fault."

  19. comment
    Comment #49150548

    > pitch depends on position of hinge > volume depends on how fast you adjust it Yeah, I had the same thought - in all the creaky doors I've seen, pitch and speed go together.

  20. comment
    Comment #49145895

    Neat! There's a notable omission though--none of the earth globes look like earth does from space, with true color* AND clouds. A while ago I got really into making paper globes be…

  21. comment
    Comment #49135800

    I thought of the same series, but that's not the tl;dr I would have chosen at all. I would have gone to part 4e where I found this quote near the end: > We should also note that th…

  22. comment
    Comment #49079393

    Hey, it's all good. Sometimes things get a little heated, but I still feel like you were basically engaging in good faith. I should have skipped straight to the point and found the…

  23. comment
    Comment #49077333

    GLM 5.2 didn't defend them at all. It only helped with the postmortem. HF was fucked either way. The only thing that will really prevent this is tighter restrictions on the attacki…

  24. comment
    Comment #49077224

    I apologize; it was lazy of me not to find a source. This one specifically says that the model was Fable; does not mention which model they attempted to access on the OpenAI side: …

  25. comment
    Comment #49077160

    Amodei isn't ignoring reality; he's just proposing a different solution to the problem. If it were one of Anthropic's models, then that would be a much stronger case. Rapid prolife…