Live data from Hacker News

Viewing profile — sozal

sozal

HN member
Joined
Tue, Oct 17, 2017, 9:21 AM UTC
HN karma
112
Public activity
60 items

About sozal

No profile information was provided.

Recent public activity

  1. story
  2. comment
    Comment #48829523

    What I'm expecting is stop spamming here with your slops, that's all.

  3. comment
    Comment #48829515

    It is not the AI slop, it is the benchmark prepared and done by humans which people like you can't understand. You still keep posting slop comments without ZERO concrete argument. …

  4. comment
    Comment #48827945

    Ok, even though "poorly written" is just your bias to me, at least you could provide some concrete critiques. It seems that you listened to the others, had the AI read the blog pos…

  5. comment
    Comment #48827865

    It's not too long for humans, but it seems that it's too long for you maybe because you don't know how to read :) You didn't finish it, so you assumed nobody could. Don't speak for…

  6. comment
    Comment #48827822

    Nice try with the metaphor, but it is just useless. I'm pretty sure that you didn't event open the link, but just read the other comments, saw the word slop and wrote something to …

  7. comment
    Comment #48824250

    @neetle I get the frustration, but reaching for "slop" as the default reaction to anything you find too long isn't all that natural. Happy to hear specific feedback on what didn't …

  8. comment
  9. comment
    Comment #48824213

    @subscribed To make a critique, first you must READ. Otherwise, your comments will be slop

  10. comment
    Comment #48824151

    Zero engagement with anything the post actually says. You've spent more words performing your distaste than it would've taken to read the thing. Recycling slop as a catch all for "…

  11. comment
  12. comment
  13. comment
    Comment #48822904

    Or just don't read the blog post and stop posting your slop here so everyone will be happy

  14. comment
    Comment #48818007

    Fair on the length. The verification loop took DeepSeek from ~20 to ~80 on Web-Bench, matching Opus at ~7x less cost. Curious if the method holds up for you.

  15. comment
  16. story
  17. comment
    Comment #47689672

    This seems fantastic! Age of Browser DevTools :)

  18. story
  19. story
  20. comment
    Comment #47352311

    Yeah, you’re right. The most of the gain is from the execute tool (batching many steps in one call), not from CDP vs Playwright. If Playwright MCP added a similar sandboxed “run sc…

  21. comment
    Comment #47351433

    We ran the same verification task with Claude Code, same prompt, 3 runs per MCP. Browser DevTools MCP: ~330K tokens, 12 turns, ~2m45s. Playwright MCP: ~1.5M tokens, 48–52 turns, ~6…

  22. story
  23. comment
    Comment #46632988

    AI can write code, but testing/debugging still needs humans. This MCP server gives AI a Playwright-powered browser with screenshots, console/network inspection, Web Vitals, ARIA/AX…

  24. story
  25. comment
    Comment #45275938

    The Driflyte MCP Server exposes tools that allow AI assistants to query and retrieve topic-specific knowledge from recursively crawled and indexed web pages. With this MCP server, …