Live data from Hacker News

Viewing profile — hollowpython

hollowpython

HN member
Joined
Wed, Jan 12, 2022, 4:46 PM UTC
HN karma
34
Public activity
42 items

About hollowpython

No profile information was provided.

Recent public activity

  1. comment
    Comment #41233272

    No wonder goths expanded so much. They have very cheap infantry.

  2. comment
    Comment #41223006

    Did they test whether GPT4 replicated already existing social science experiments? If so, this might have happened because the experiment was in the training data

  3. comment
    Comment #41065373

    I'm a programmer and data scientist. Using LLMs (specifically Claude 3.5 Sonnet) made me at least 10x as productive. I don't write boilerplate code anymore.

  4. comment
    Comment #41058979

    > The hard thing about this I see what you did there.

  5. comment
    Comment #40451654

    Do you really do this, and is striking out a line of a contract binding?

  6. comment
    Comment #40104102

    Serious answer: do you want me to build it for you? If you give me a contact email, we can discuss the budget

  7. story
    How to generate consistent faces/characters with LLMs?

    I want to use LLMs to generate consistent characters (with consistent faces) for rpg games. Does anyone know of a way to do this programatically? Either via API or offline. Thank y…

  8. comment
    Comment #39976761

    Are there any which can generate consistent characters (especially faces)?

  9. comment
    Comment #38909607

    Winning rock-paper-scissors enough times against cgpgray: https://m.youtube.com/watch?v=PmWQmZXYd74

  10. comment
    Comment #38719570

    To better understand the validity of this document, check this wikipedia page: https://en.wikipedia.org/wiki/Necessity_and_Urgency_Decree . I used google translate to produce an En…

  11. story
  12. comment
    Comment #37584179

    Does anyone know a tool like this but for arbitrary PDFs?

  13. comment
    Comment #36859682

    Well, everyone knows Mongol cavalry archers fire 25% faster...

  14. comment
    Comment #36537791

    I agree that the feedback loop wouldn't make the models smarter than us, but training data would. I can see how with more computing power we will be able to feed it the whole of Yo…

  15. comment
    Comment #36432585

    Try eating more! I have a similar lifestyle to you, and I discovered I was undereating. When I increased calories by 20%, suddenly my memory wasn't so foggy, I was more motivated..…

  16. story
    Ask HN: Data Science consultants, how did you start?

    I'm looking to transition from Bug Tech IC to consultant. Does anyone want to share their experiences? For context, I'm in Germany. I'm also interested in what kind of projects you…

  17. comment
    Comment #33246255

    Does it use any kind of recommendation engine based on my current tastes/already liked songs?

  18. comment
    Comment #33246244

    Oh now it works! But it wasn't intuitive for me that I needed to do that

  19. comment
    Comment #33153091

    I logged in but it doesn't show me any songs :(

  20. story
    Ask HN: Is astrological data used in ML models?

    A positive argument for astrology being true would be that a person's sign, natal chart or characteristics derived from it are used as attributes for machine learning models. Is th…

  21. comment
    Comment #32992192

    About #10: issn't undecidability the same as uncomputability? AFAIK, they are synonyms.

  22. comment
    Comment #32493144

    I think it's easier to get a new job if you have a job, otherwise they can low-ball you because you look more desperate

  23. comment
    Comment #32366177

    I'm confused, why do people need to miss the feature? If it's that important to them, just don't buy the newest iphone.

  24. story
    Ask HN: Do you use GPT-3 like models at work? How do you validate their output?

    Hi! I work in NLP. I'm curious about finding use cases for GPT-3 like models, but an issue I have is that these models sometimes produce garbage, which is dangerous. How do you do …

  25. comment
    Comment #32317570

    Ooooh. The prisoners can't hear other prisoners' guesses then. Ok! But in your example, isn't the probability of winning 1-(99/100)*100? I don't think your "best case scenario" com…