Live data from Hacker News

Viewing profile — powera

powera

HN member
Joined
Sat, Oct 23, 2010, 6:58 AM UTC
HN karma
2,395
Public activity
962 items

About powera

Formerly of Google and Quip.

Recent public activity

  1. comment
    Comment #48508386

    The priesthood doesn’t like that the peasants can read the Bible for themselves now.

  2. comment
    Comment #48388376

    I'm seeing very low quality results on LMStudio with this model. Worse than Gemma 3 12B. It is getting questions like "David has 18 apples and Ivan has 7 apples. How many apples do…

  3. comment
    Comment #47911961

    He (or ChatGPT) is throwing spaghetti at the wall. Not having the standard API key be able to delete the database (and backups) in one call makes sense. "Wanting a human to type DE…

  4. comment
    Comment #47891613

    I'm not sure they've found/understand it yet. My two main theories: 1. A bunch of people with new Claude Code codebases in December now are working with a larger codebase, causing …

  5. comment
    Comment #47739732

    From March, also https://blog.fontawesome.com/we-have-a-99-email-reputation-g... is the canonical URL.

  6. story
  7. story
  8. comment
    Comment #47711011

    Click-bait title; the article goes on to say "Open source isn't actually broken" as long as you buy their product.

  9. story
  10. story
  11. comment
    Comment #47586748

    Wikinews never worked; the principle of "verifiability" that Wikipedia was based on simply doesn't work for news-collection, which requires trusted first-party accounts. The projec…

  12. comment
    Comment #47489822

    (January 2025)

  13. comment
    Comment #47417085

    So far on my (simple) benchmarks, GPT-5.4-mini is looking very good. GPT-5.4-mini is about 30% faster than GPT-5-mini. GPT-5.4-mini gets 80% on the "how many Rs in Strawberry" test…

  14. comment
    Comment #47415707

    I've been waiting for this update. For many "simple" LLM tasks, GPT-5-mini was sufficient 99% of the time. Hopefully these models will do even more and closer to 100% accuracy. The…

  15. story
  16. comment
    Comment #47219336

    This looks like somebody re-releasing QWEN models to promote their own company. https://news.ycombinator.com/item?id=47217305 is the link to QWEN's repo.

  17. story
    Claude Code on the Web broken?

    I have been trying to use Claude Code on the Web (and also mobile) this morning, and the LLM responses are never showing up. There is a diff generated (so something is happening), …

  18. story
    Ask HN: Best multi-lingual text-to-speech system

    I'm looking for a way to bulk-generate audio based on text files. Ideally, it would be a system I can run locally (M3 mac, 24GB RAM), and support at least 10 languages natively. I …

  19. story
  20. story
  21. comment
    Comment #47035134

    Between this and 4.6's tendency to do so much more "exploratory" work, I am back to using ChatGPT Codex for some tasks. Two months ago, Claude was great for "here is a specific tas…

  22. comment
    Comment #47004227

    I've been working on a similar app called Trakaido. (Yes, it's also AI generated content). The demo leans heavily on "choose the words for the sentence", which avoids spelling/keyb…

  23. story
    What Is Genspark?

    One of the Super Bowl commercials today was from Genspark.ai , a company I had not heard of before today. Their website looks like a generic ChatGPT clone. Their LinkedIn page boas…

  24. comment
    Comment #46913046

    There's a difference between the chatbot "advertising" something and an hour-long manipulative conversation getting the chatbot to make up a fake discount code. Based on the OP's c…

  25. comment
    Comment #46900849

    This seems to have a fundamental misunderstanding of Hacker News's front page. The goal is to have posts only show up for at most a day or two. I would expect 100% to be gone in a …