Preferences

ayewo
Joined 1,150 karma
https://twitter.com/ayewo_

  1. Yes, you are not imagining things :)

    Discussed on HN 3 months ago: https://www.hackerneue.com/item?id=44972151

  2. Would love to see it when you do get round to posting it :)
  3. Terminology-wise, does this read like a better title instead?:

    Show HN: Gemini Pro 3 generates the HN front page 10 years from now

  4. Perhaps fabricate is a better word, as suggested elsewhere in this thread?
  5. Like whitewashing, but for AI, I’m guessing.
  6. That 1:19 clip was quite good actually. Thanks for the laugh :)
  7. > Set me back 2 weeks of work.

    How did this happen?

    Did you let the agent loose without first creating its own git worktree?

  8. Is the code to your agent and its implementation of "str_replace_based_edit_tool" public anywhere? If not, can you share it in a Gist?
  9. You are ignoring the possibility of technological disruption.

    Apple disrupted Nokia and Blackberry. ARM is currently disrupting Intel.

    What if someone lands on a break-through using a completely different tech: what if X-ray lithography [1] becomes viable enough that they don’t have to acquire state-of-art EUV machines from ASML?

    [1] X-ray lithography was abandoned in the 80s but it is being revisited by Substrate https://substrate.com/our-purpose. They are an American company that hopes to make it commercially viable by being cheaper and far less complex than EUV.

  10. Congrats on shipping!

    Apart from the obvious question of why you didn't opt to open source the tool :), I'm genuinely curious about how you approached development.

    How did you decide for this feature A: "I'll just spawn child processes and read the output of `x-y-z` and `a-b-c` CLI tools, while for feature B: "I'll drop-down to BSD sockets"? Perhaps you have a performance budget: if using the Apple-provided CLI utilities are not fast enough then you drop down to writing BSD sockets?

  11. > The next logical step would be to get into CPUs, to become a fully integrated computing solutions provider.

    They already tried it 5 years ago [1][2] but it was promptly blocked by regulators.

    [1]: https://nvidianews.nvidia.com/news/nvidia-to-acquire-arm-for...

    [2]: HN discussion https://www.hackerneue.com/item?id=24464807

  12. Services were rendered for just 4 hours is somehow worth $30k?
  13. I’m curious: how did you go about the implementation? You built an MCP server for D2?
  14. This recent comment https://www.hackerneue.com/item?id=45598670 by @simonw really helped drive home the point that LLMs are really being fed an array of strings.
  15. An open source model does exist now [1] and is multilingual. Previous discussion [2].

    [1] https://ethz.ch/en/news-and-events/eth-news/news/2025/07/a-l...

    [2] https://www.hackerneue.com/item?id=44535637

  16. To add to this, modules make it easy for external tools like GraalVM native-image to produce self-contained binaries that are smaller compared to the standard practice of distributing fat binaries (i.e. large JARs).
  17. There's a fork of the repo you posted: https://github.com/QuanZhang-William/gemini-cli-security that's listed on Google's extensions gallery page: https://geminicli.com/extensions/browse/

    EDIT:

    I've posted about it on GitHub: https://github.com/gemini-cli-extensions/security/issues/81

    Hopefully the relevant team will see it there.

  18. "Signups are open [1], and will remain so until October 20th."

    Recently on HN [2].

    [1]: https://solve.it.com/

    [2]: https://www.hackerneue.com/item?id=45455719

This user hasn’t submitted anything.