Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Show HN: Jev-windows-agent – Windows UI Automation back end for CUA agents (github.com/vbs2004)
    —discuss
  2. Being Mentioned in AI Answers Is Not the Same as Being Recommended (cuescout.com)
    —discuss
  3. SRE Udemy Course: Software Reliability Map (udemy.com)
    —discuss
  4. Jempo by Thomas Ha (lightspeedmagazine.com)
    —discuss
  5. "As a Language Model": Chat Template Switches LLM Self-Referential Voice (arxiv.org)
    —discuss
  6. Changes in quality of life: digital exercise therapy for knee/hip osteoarthritis (doi.org)
    —discuss
  7. Postgres AT TIME ZONE 'UTC' does NOT do what you think it does (bookofrevenue.com)
    —discuss
  8. Trying GPT-6 Sol with AI SDK Evaluation API (vercel.com)
    —discuss
  9. Show HN: TanStack Start and TanStack AI Jev Integration (vercel.com)
    —discuss
  10. Another week, another data breach for Revolut customers (theregister.com)
    —discuss
  11. Ask HN: What does your AI workflow look like?
    1comments
  12. The Importance of Human Knowledge in the AI Era (jonbehnken.substack.com)
    —discuss
  13. Claude Isn't Allowed to Write Me Prose (kvit.app)
    —discuss
  14. YouTube stores three auto-extracted frames for every video, and nobody uses them (github.com/infinityloop1308)
    —discuss
  15. How AI Is Accelerating PCB Design and Prototyping (eetimes.com)
    —discuss
  16. Poisoning a Go Cache (lukasschwab.me)
    1comments
  17. Genius AI APı Anahtarı
    —discuss
  18. Makeraoke: Share your side project in a video between three and nine minutes (reddit.com)
    —discuss
  19. Musk – Official Trailer (2026) (youtube.com)
    —discuss
  20. Sandbox-first AI coding harness (chock.ws)
    —discuss
  21. Show HN: Peekaboolean – image and jev-like typed questions in typed anwers out (github.com/bykof)
    —discuss
  22. I Posed as a Problem Gambler. DraftKings Made Me a VIP (propublica.org)
    1comments
  23. Show HN: Afterlife Vault – a virtual dead man's switch (afterlife-vault.com)
    1comments
  24. Using Claude Code: Spending your effort (claude.dev)
    —discuss
  25. The year AGI happened, and some unhinged predictions (ouzu.im)
    —discuss
  26. Show HN: LightCloud – A cloud console organised like file system (light-cloud.com)
    —discuss
  27. From game physics to care robots (2023) [The Royal Society] [video] (youtube.com)
    —discuss
  28. How can you quikly tell if a technical article lacks firsthand experience?
    —discuss
  29. There are no borders or boundaries on our planet (2023) (stargazingguy.co.uk)
    —discuss
  30. Rust for C# and .NET developers guide by Microsoft (microsoft.github.io)
    —discuss

Ask HN: Is answering GitHub issues with AI without declaring it a big deal?

1 pointsby 1h ago
1 comments
I'm currently using the Mindwtr app to manage tasks, it's great and cross platform and all.

However I wanted to propose a feature and I looked at the issues and got a bit weirded out by the tone, and it's clear to me they are using an AI to write their responses, but even more intesting it seems also fully automated.

Things like saying to someone having the contributor tag on GitHub (and who also seems writing using an LLM) "Great breakdown — you've clearly looked at the codebase." after they provided input on a feature

You can see the thread here and many more examples in the repo: https://github.com/dongdongbh/Mindwtr/issues/361

Right now I'm a bit torn, the app works well, it's open source, but at the same time I don't feel like it's okay to do this, I see it a bit like abandoning a project to the whims of the LLM.

So, would you keep using this app or not ? And why ?

58m agoHN ↗

Give credits to AI where credits is due while also mention the human effort is