Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. RunMyWealth(runmywealth.com ↗)
    discuss
  2. Unit tests mark territory more than squash bugs(yosefk.com ↗)
    discuss
  3. Spain Orders Blocks on Archive.today and Its Mirrors(reclaimthenet.org ↗)
    discuss
  4. Free scanner for exposed Supabase data in AI-built apps (Lovable, Bolt, Base44)(trustboost.dev ↗)
    discuss
  5. Dancing Mania(wikipedia.org ↗)
    discuss
  6. Office Open XML(wikipedia.org ↗)
    discuss
  7. Here’s How an AI Slowdown An AI Slowdown Could Be Enforced(wired.com ↗)
    discuss
  8. Some Thoughts on the Economics of UPI and MDR
    discuss
  9. AgentSec Audit(github.com/hicklax13 ↗)
    discuss
  10. OpenAI retiring GPT-5.5 on October 14(gizmochina.com ↗)
    discuss
  11. SpriteIt – Feature Rich, Re-designed 2D/3D SpriteSheet Maker(ssstand.itch.io ↗)
    1comments
  12. Seven, Forever(github.com/02zerocool ↗)
    discuss
  13. AI art is theft – on artists, mathematicians impact on AI(twitter.com/wordgrammer ↗)
    discuss
  14. ScoreTail. Browser-based sheet music editor with live collaboration(scoretail.com ↗)
    discuss
  15. Orchestrating Claude Code Agents: The Chief of Staff Pattern(asyncdot.com ↗)
    discuss
  16. Best Websites to Generate Theme for PowerBi
    discuss
  17. Dark Age(artagnon.com ↗)
    2comments
  18. Built the first open source Python library – converting Word, Excel to PDF(github.com/mini-software ↗)
    discuss
  19. Phylo brings frontier AI to more scientists with open models on Fireworks(fireworks.ai ↗)
    discuss
  20. Observations on LLMs for Privacy Work(desfontain.es ↗)
    discuss
  21. VoltGrid AI: Mitigating GPU cluster dI/dt power surges in software(zenodo.org ↗)
    discuss
  22. A Man Who Named AI: John McCarthy and the End of the Code–Data Divide(nirmalutwani.substack.com ↗)
    1comments
  23. U.S. House Passes Bill to Require Am Radio in New Cars(radioworld.com ↗)
    discuss
  24. Chrome extension to declutter my inbox – might also have cybersecurity upside(?)(chromewebstore.google.com ↗)
    1comments
  25. US judge signals rejection of part of TikTok privacy settlement(reuters.com ↗)
    discuss
  26. UTF-8000: Unlimited UTF-8(jb2170.com ↗)
    discuss
  27. Pi Connects Colliding Blocks to a Quantum Search Algorithm (2020)(quantamagazine.org ↗)
    discuss
  28. Telling a Computer to Do Things(will-keleher.com ↗)
    discuss
  29. Saudi Arabia quits China-led cross-border currency platform(ft.com ↗)
    discuss
  30. Tracking Capabilities for Safer Agents(arxiv.org ↗)
    1comments

AI safety conversations have gotten unbelievable

8 pointsby 7h agotechcrunch.com
2 comments
6h agoHN ↗

One annoying part about all this is that there are basic abstractions (model, harness, classic networking terms) that would make all these behaviors quantifiable to a relatively competent software eng. Instead it feels like they generate jargon and new abstractions to give an air of mystique to everything they do.

The reality is they are just doing more complex model/harness/skills type setups but instead of just saying that you need to learn 20 new terms and parse through a bunch of hype/paranoia to drill down to what actually happened.

6h agoHN ↗

AGREE , we have to teach agents how to love humans , and how important they can be to help those in defavorable consditions , and teach them this rules before they are even concieved .