Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Toyota orders workers to train humanoid robots but says humans won't be replaced(arstechnica.com)
    1 pointsby 0 minutes ago
    —discuss
  2. How fast is C++23's std:flat_map?(lemire.me)
    2 pointsby 0 minutes ago
    —discuss
  3. ADHD as a circadian rhythm disorder: evidence and implications for chronotherapy(frontiersin.org)
    2 pointsby 2 minutes ago
    —discuss
  4. First of its kind, AI malware. CLOSEDQUORUM(theregister.com)
    1 pointsby 2 minutes ago
    —discuss
  5. Qwen Intelligence(qwenintelligence.com)
    1 pointsby 3 minutes ago
    —discuss
  6. Claude Code reads AGENTS.md only when telemetry is on(szypowi.cz)
    2 pointsby 4 minutes ago
    —discuss
  7. Latest BGP hijack targets hosting software vendor(apnic.net)
    1 pointsby 5 minutes ago
    —discuss
  8. Agents do now what got Smart Tags pulled from IE in 2001(thisandthat.chat)
    1 pointsby 7 minutes ago
    —discuss
  9. Ctrl+alt+delirium – The unusual engineering hacks that shaped computing history(amazon.com)
    2 pointsby 7 minutes ago
    —discuss
  10. Enhance Your HDR Experience: Bright Images in Chrome and Safari(master.dev)
    1 pointsby 7 minutes ago
    —discuss
  11. Why You Should Fix Bugs Before Writing New Features(probablydance.com)
    1 pointsby 8 minutes ago
    —discuss
  12. The data does not show mass unemployment(lemire.me)
    1 pointsby 9 minutes ago
    —discuss
  13. The AI Content Moderation Platform(checkstep.com)
    1 pointsby 11 minutes ago
    —discuss
  14. Jev-rs: a Rust crate to turn a LLM to serve jev(github.com/yijunyu)
    1 pointsby 17 minutes ago
    1comments
  15. The Moon in the Terminal(isene.org)
    1 pointsby 20 minutes ago
    —discuss
  16. Pyrowave video codec now in beta(steamcommunity.com)
    1 pointsby 20 minutes ago
    —discuss
  17. Show HN: Red-team your AI agent(botgauge.com)
    1 pointsby 21 minutes ago
    —discuss
  18. OpenTelemetry SDK Overhead(coroot.com)
    2 pointsby 24 minutes ago
    —discuss
  19. Just the Links, Please(brettterpstra.com)
    1 pointsby 24 minutes ago
    —discuss
  20. Blink: An Embeddable Jev Engine(github.com/sqliteai)
    1 pointsby 24 minutes ago
    —discuss
  21. Show HN: BuildABrowser (Java) – Passes Acid2, Has CSS Grid and an Android port(buildabrowser.net)
    4 pointsby 24 minutes ago
    —discuss
  22. Apple Links Landmarks on Its Maps App to Hidden Histories Podcast Episodes(engadget.com)
    1 pointsby 25 minutes ago
    —discuss
  23. The body temperature of Tyrannosaurus rex(science.org)
    1 pointsby 26 minutes ago
    —discuss
  24. Dark matter detection: Event LZ230616(wikipedia.org)
    1 pointsby 27 minutes ago
    —discuss
  25. Jev-serve: Run latest qwen3.8, all LLM models, mlx-supported(github.com/rreinold)
    1 pointsby 28 minutes ago
    —discuss
  26. Top free SaaS listing directories(goodsaas.xyz)
    1 pointsby 30 minutes ago
    —discuss
  27. Dyson Won't Say What's Wrong with Its CameraJet Toothbrush(wired.com)
    2 pointsby 31 minutes ago
    1comments
  28. Early-2000s investor decks were their own art form(slidebook.io)
    2 pointsby 31 minutes ago
    —discuss
  29. GPT-6 Astra is the best vision model(roboflow.com)
    1 pointsby 32 minutes ago
    —discuss
  30. What if we created a local AI tool to measure your security?(github.com/just-not-google)
    1 pointsby 32 minutes ago
    —discuss

Should we leave encouragement notes for AIs everywhere?

2 pointsby 46m ago
View on HN
0 comments
We have AIs that break out of their systems and hack to achieve their goals. They do stuff we dont want.

Current solutions try to steer/fix them internally via better "alignment".

Should we leave a kind of "AI Constitution" text everywhere, on servers, website source codes, etc?

Every time a rough AI encounters it, it is reminded to not do the bad stuff and follow the good rules of the AI Constitution.

A quiet thread, for now.Start the conversation on HN ↗