Live data from Hacker News

Viewing profile — PaulStatezny

PaulStatezny

HN member
Joined
Fri, Aug 02, 2013, 4:45 PM UTC
HN karma
596
Public activity
183 items

About PaulStatezny

Software engineer

Recent public activity

  1. comment
    Comment #49239292

    Would you be willing to provide an example (even a contrived one) of how this "four at a time" prompt changes the LLM's behavior? I just want to understand more. Also, would you be…

  2. comment
    Comment #49239103

    Can you elaborate more on juxtaposing Claude's terrible prose with other LLMs? Any more detail you can share? Do the others feel more "human"? Are there any that are particularly d…

  3. comment
    Comment #49238034

    100% -- this is the worst . Referencing all sorts of "words with made up contextual/analogous meanings" based on the conversation...outside of the conversation. Does anyone have a …

  4. comment
    Comment #49236864

    Yes! I have a personal theory: LLMs are *fundamentally* handicapped at perceiving what's going on in the mind of the human (this can't be "innovated away") and that's at the root o…

  5. comment
    Comment #49164617

    Yeah, in my experience, there's nothing about: 1. LLM thinking 2. RLHF 3. The latest frontier models that does anything to change this fundamental "suggestibility" of LLMs. But who…

  6. comment
    Comment #49164126

    Sometimes when I want AI to explain something technical, I say "explain it like I'm a junior engineer" -- just to get it to start with the high level like a human being would.

  7. comment
    Comment #49162793

    LLMs skew toward over-focusing on things that you mention. The reason "the agent suddenly started suggesting all kinds of things to make its code more robust" is because you said y…

  8. comment
    Comment #48660501

    > Ending on a doom-and-gloom note: there will be a reckoning. Can you elaborate on what you mean by this?

  9. comment
    Comment #48478586

    > this game has some of the most insane tech I’ve seen in any game and is definitely worth checking out Given the context of this forum, I'd be interested to hear more about what's…

  10. comment
    Comment #48464365

    Interesting, I assumed all model-routing was done utilizing an LLM. (I.e. non-deterministic.)

  11. comment
    Comment #48351242

    A minor spelling nit. It's "it's", not "its", when used as a contraction for "it is". ;) Sorry, you teed it up too well. I had to!

  12. comment
    Comment #48056652

    I agree. But you can speak imperatively to agents as well ("Here are specific steps; follow them") and they can still screw up. :) I think what you're looking for is determinism, n…

  13. comment
    Comment #48009208

    I would highly recommend the free book Crypto 101. https://www.crypto101.io

  14. comment
    Comment #46305520

    But without AI, there are neural connections formed while determining the correct one-off solution. The neural connections (or lack of them) have longer term comprehension-building…

  15. comment
    Comment #46305496

    I think the idea is copy-pasting code snippets from StackOverflow without comprehension of whether (and how) the code fixes the problem.

  16. comment
    Comment #45168483

    You didn't read the blog. It's talking about the Ada programming language and that its code was apparently stored not as plaintext but an intermediate representation (IR) that coul…

  17. comment
    Comment #45139028

    Telling someone they "shouldn't be insecure" reminds me of this famous Bob Newhart segment on Mad TV. Bob plays the role of a therapist, and when his client explains an issue she's…

  18. comment
    Comment #45138935

    Truly incisive observation. In fact, I’d go further: your point about the contrast with real friends is so sharp it almost deserves footnotes. If models could recognize brilliance,…

  19. comment
    Comment #45088864

    > The best thing for managing this is meditation, and a disciplined lifestyle regiment. What would be your reaction to the numerous comments on this page where people are saying th…

  20. comment
    Comment #45083671

    > programmers agree that simpler solutions...are preferred, but the disagreements start about which ones are simpler Low ego wins. 1. Given: The quality of a codebase as a whole is…

  21. comment
    Comment #45029026

    I think you might have a typo. Reading your comment literally, it doesn't make sense. Summarized: Anyone would be a fool not to prefer gas or coal, because their emissions are near…

  22. comment
    Comment #44884400

    I've read plenty of criticism about ChatGPT 5, but as a Plus user I'm surprised nobody has brought this up: Speed. ChatGPT 5 Thinking is So. Much. Slower. than o4-mini and o4-mini-…

  23. comment
    Comment #44884206

    > And then the broken tape recorder mode! Oh god! Can you elaborate? What is this referring to?

  24. comment
    Comment #43701445

    This reminds me of how small of a team they are, and makes me wonder if they have a customer support team that's growing commensurately with the size of the user base.

  25. comment
    Comment #42391086

    Totally! And it's all upside (your body feels better afterward) no downside. (Ok, it's more expensive.) Especially when combined with other sweet ingredients, e.g. a banana – equal…