Live data from Hacker News

Viewing profile — Sunhold

Sunhold

HN member
Joined
Mon, Jan 31, 2022, 4:32 PM UTC
HN karma
161
Public activity
52 items

About Sunhold

No profile information was provided.

Recent public activity

  1. comment
    Comment #41657829

    It shows that the LLM is capable of reasoning.

  2. comment
    Comment #41657717

    Look at the sample chain-of-thought for o1-preview under this blog post, for decoding "oyekaijzdf aaptcg suaokybhai ouow aqht mynznvaatzacdfoulxxz". At this point, I think the "fan…

  3. comment
    Comment #40398590

    I now write most of my personal scripts in English and have ChatGPT compile it to whatever programming language. I don't use it for anything too complex but it makes it easier to w…

  4. comment
    Comment #40210718

    Works for me https://chat.openai.com/share/51c55c5e-9bb2-4b8c-afee-87c032...

  5. comment
    Comment #39725054

    I tried with GPT-4 and it can't recognize the font shown in the article without context.[1] (You can generate it as as the 'block' font using FIGlet or the many online generators t…

  6. comment
    Comment #38868049

    https://annas-blog.org/how-to-run-a-shadow-library.html

  7. comment
    Comment #38753888

    This guy seems to think every problem means civilization is going to collapse. Microplastics cause cancer, therefore civilization is doomed. Meanwhile back in reality, cancer morta…

  8. comment
    Comment #38590697

    The custom GPTs are convenient. Instead of digging through my collection of prompts for different tasks, I just click one of my custom GPTs that has been preinstructed.

  9. comment
    Comment #38360466

    I would rather OpenAI have a diverse base of income from commercialization of its products than depend on "donations" from a couple ultrarich individuals or corporations. GPT-4 cos…

  10. comment
    Comment #38351023

    Why should they trust the board? As the letter says, "Despite many requests for specific facts for your allegations, you have never provided any written evidence." If Altman took a…

  11. comment
    Comment #38331801

    She cofounded Fellow Robots, which shipped a product scanner on wheels to 11 stores[1] and then quietly folded up (homepage is now parked by GoDaddy). [1] https://www.therobotrepor…

  12. comment
    Comment #38266628

    It always amazes me that anyone can think this comic makes a good point.

  13. comment
    Comment #38141383

    Every time I use it I am irritated that it won't let me copy the name of my graphics card or other components and I have to retype them manually.

  14. comment
  15. comment
    Comment #37914320

    This does not seem like a good faith engagement. I scrolled to a random page and it said, "Take the claim in the History of the Jews in Poland article, that Jews have specific phys…

  16. comment
    Comment #37880327

    Note that it only recommended that because they intentionally prompted for it.

  17. comment
    Comment #37459778

    The article discusses this in detail.

  18. comment
    Comment #37459362

    I don't know why you are assuming it is standard, expected or necessary to have genome sequencing equipment on the ship. There's no doubt that they've brought it back to land and a…

  19. comment
    Comment #37335932

    Have you never left a city? The idea that half the forests in Canada are plantations should be ridiculous on its face.

  20. comment
    Comment #37306638

    My guess is that they accidentally trained it to object to the past for being racist, i.e. "once upon a time" promotes "outdated" attitudes.

  21. comment
    Comment #37175744

    It broadly prohibits "military, warfare, nuclear industries or applications, espionage, use for materials or activities that are subject to the International Traffic Arms Regulatio…

  22. comment
    Comment #37145596

    If it's trained on pictures of children and adult porn, the model can easily combine the two concepts. Even if you remove one of these from the training set, it can be finetuned ba…

  23. comment
    Comment #37042504

    Why do you believe you can brute-force cache links? There could be some vulnerability but that string looks far too long to brute-force.

  24. comment
    Comment #36892669

    Sam Altman has never denied that GPT-4 is a mixture of experts model. He denied an early rumor that it was a 100 trillion parameter model.[1] The mixture of experts rumor states th…

  25. comment
    Comment #36802035

    This is pure speculation. WhatsApp is banned in some authoritarian countries such as China and UAE. It probably remains unbanned in others simply because it is so important for bus…