Live data from Hacker News

Viewing profile — guluarte

guluarte

HN member
Joined
Sat, Feb 11, 2023, 8:11 PM UTC
HN karma
335
Public activity
396 items

About guluarte

No profile information was provided.

Recent public activity

  1. comment
    Comment #49258703

    will this solve anything? people will build tools to remove the watermarks

  2. comment
    Comment #49249371

    the problem with this is... now i trust the agents more than myself lol

  3. comment
    Comment #49182708

    cli tools > mcps

  4. comment
    Comment #49174512

    I think most of the cost is in R&D

  5. comment
    Comment #48960360

    i just use terminus + wireguard + tmux and works great, i can control claude/codex from my phone while working out

  6. comment
    Comment #48951152

    cool, meanwhile people will use pixelated pepes instead

  7. comment
    Comment #48925380

    "Hey Codex, help me design the most useless hardware you can think of"

  8. comment
    Comment #48897823

    2 base jumps? are those really that dangerous?

  9. comment
    Comment #48886840

    Check all MCP you have enabled in claude.ia

  10. comment
    Comment #48848770

    They announced that GPT-5.6 Sol, along with Terra and Luna, will launch publicly today.

  11. comment
    Comment #48846841

    A lot of these benchmarks are unfamiliar. Are labs just choosing the ones that make them look best?

  12. comment
  13. comment
    Comment #48826036

    I don't think one week is enough to learn the complex business rules that some software needs to follow.

  14. comment
    Comment #48819116

    At the moment, AI is creating more work for me, not less. The more capable the models become, the more responsibilities and expectations you have as a developer. Nowadays, I spend …

  15. comment
    Comment #48809513

    At least they’re honest and not blaming AI.

  16. comment
    Comment #48766675

    use /goal with bypass permission then

  17. comment
    Comment #48766656

    for me fable does things nobody asked for

  18. comment
    Comment #48720010

    This feels like telecom "massive demand, bad returns". If a good enough model can be swapped in every few months, the value moves away from the model and toward cheap inference. Th…

  19. comment
    Comment #48668083

    or tell an llm to do it for you

  20. comment
    Comment #48667937

    Anthropic training their models full of copyright data, so?

  21. comment
    Comment #48571035

    I would agree but in this case they are buying the talent and the data.

  22. comment
    Comment #48512781

    Also fable was good but not Manhattan level project, i honestly did not find a major difference between it and gpt 5.5

  23. comment
    Comment #48512655

    well... Leason learned i guess, they hyped mythos too much

  24. comment
    Comment #48491327

    For me, AI can sometimes create a false sense of productivity. It's similar to how in the past, people would spend time creating the perfect setup with notion templates, pomodoro t…

  25. comment
    Comment #48469175

    AI creates more work