Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. There's a high chance of devices being sold with GrapheneOS preinstalled in 2027(grapheneos.social)
    discuss
  2. Moving from cash to credit cards, PayPal, etc. is an ongoing privacy disaster(grapheneos.social)
    discuss
  3. In 200-Page Report, Cornell Confronts the Crisis in American Higher Education(wsj.com)
    discuss
  4. Shall We Repeal the Laws of Economics – Part III(oaktreecapital.com)
    discuss
  5. Alzheimer's Is No Longer an Untreatable Disease(sciencealert.com)
    discuss
  6. A golden opportunity: Seattle's surveillance pricing ban(thenexusofprivacy.net)
    discuss
  7. ASML Executive Says It Has No Sales in Europe(bloomberg.com)
    discuss
  8. Bugcrowd is currently fundamentally broken(leonbecker.de)
    discuss
  9. Why is social media so humourless?(baldurbjarnason.com)
    discuss
  10. Ask HN: How do your teams share and distribute agent skills?
    discuss
  11. What drives BigQuery costs, and what doesn't(erathos.com)
    1comments
  12. Should TypeScript support runtime types instead of relying on Zod?(twitter.com/mykhailen)
    1comments
  13. We are the last generation of human psychiatrists(cambridge.org)
    discuss
  14. Grok bot is now in Tesla(twitter.com/elonmusk)
    1comments
  15. Implementing the Camera Mechanic from Viewfinder(ishamf.dev)
    discuss
  16. Show HN: Ttmux, a Modern and Fast Tmux(github.com/statico)
    1comments
  17. Anthropic launches Claude Opus 5.5 with stricter safeguards for cybersecurity(theverge.com)
    discuss
  18. Claude Opus 5.5 for code review: More catches, different misses(coderabbit.ai)
    discuss
  19. Claude Opus 5.5(claude.com)
    discuss
  20. Anthropic now offering limit resets like Codex(twitter.com/claudedevs)
    discuss
  21. Roundtables: The Deadly Failures of the Virtual Border Wall(technologyreview.com)
    discuss
  22. The Myth of AI Doom, with Cal Newport(youtube.com)
    discuss
  23. I open-sourced a small library for drawing SVGs progressively(npmjs.com)
    discuss
  24. I run a production AI app for €60 a month. Here is the actual bill.(tailstory-app.com)
    discuss
  25. Show HN: WebMCP Registry of Websites That Agents Can Use(wmcp.ai)
    discuss
  26. Show HN: Groupicorn directory of IOP group therapy programs
    discuss
  27. Ancient DNA reveals pervasive directional selection across West Eurasia(nature.com)
    discuss
  28. Code similarity detection using Tree-sitter(github.com/dsummersl)
    1comments
  29. ES Archive – a new memory system for AI, running natively on Mac(github.com/apocryphx)
    discuss
  30. Claude Opus 5.5 Intelligence, Performance and Price Analysis(artificialanalysis.ai)
    discuss

How much of the "AI is going rogue" talks is just a publicity stunt?

4 pointsby 1h ago
3 comments
1h agoHN ↗

Probably most of it. Have "AIs" invented new techniques, or are they just applying techniques humans invented? It's the latter.

Remember that first time you heard about speculative execution attacks (Spectre)? That was something new, something unheard of. If an "AI" is going rogue, something previously unheard of would show up. Otherwise, they're just doing what patterns are in their training data, and doing a lot of it. And, it's a human telling them to do so.

1h agoHN ↗

A swarm of AI agents that knows what humans know about hacking but can do it far faster... that's still a formidable attacker. Not novel. Not "escaping". But formidable.

"Escaping" would be using the attack to install a copy of itself somewhere. But we've seen that before, too, with viruses and worms. And if it has to install an 8 GB image to install a copy of itself, it's not going to be the stealthiest of worms...

1h agoHN ↗

Well, since all I've seen is publicity, I'd have to say 100%.