Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. OpenAI: Tomorrow we are re-opening the Pro $200 subscription (twitter.com/thsottiaux)
    —discuss
  2. Study: A lack of professional AI guidelines leads to feelings of "AI shame" (niemanlab.org)
    —discuss
  3. Show HN: Prompt Anonymization and Deanonymization (dataanonymiser.com)
    —discuss
  4. US Forces Exit Iraq (reuters.com)
    —discuss
  5. Web Time Lapse (projectnothing.ai)
    —discuss
  6. When can we say AI made a scientific discovery? (technologyreview.com)
    —discuss
  7. I built a small workflow execution engine around a deliberately narrow problem (github.com/varun-51)
    —discuss
  8. Show HN: Pic2svg – convert PNG/JPG to editable SVG paths in the browser (pic2svg.com)
    —discuss
  9. RuneBench: Agent Benchmark on RuneScape Gameplay Tasks (maxbittker.github.io)
    —discuss
  10. Towards safety cases for frontier AI training (openai.com)
    —discuss
  11. As AI models go rogue, do you still trust OpenAI and Anthropic to stop them? (theguardian.com)
    —discuss
  12. Trump asked Xi Jinping if China would like to buy US weapons (theguardian.com)
    1comments
  13. Mathematical Origami (mathigon.org)
    —discuss
  14. The Quantum Atlas (quantum physics for non-experts) (umd.edu)
    —discuss
  15. Britain overtaxes well-paid workers (economist.com)
    2comments
  16. Education Department formally rescinds Title IX protections for LGBTQ students (thehill.com)
    —discuss
  17. Automating eval design and hillclimbing with Claude (claude.dev)
    —discuss
  18. Humans Are Reading Copilot Prompts – and They're Horrified (404media.co)
    —discuss
  19. An Open-Source Terminal UI for Azure DevOps
    1comments
  20. What's Behind the AI Panic (bombthrower.com)
    2comments
  21. The dataset where fine-tuning turned actively harmful – and how we now catch it (cacheverifier.com)
    —discuss
  22. Hackers Jailbreak AI Chatbots (twitter.com/hamzaonchain)
    —discuss
  23. CVE-2026-85706 (CVSS 10.0) (twitter.com/eqstlab)
    1comments
  24. Suicides in 5 Weeks. What's Going on at Cyber Command at Fort Meade? (thebanner.com)
    1comments
  25. What 1,350 Runs Taught Us About Prompt Guardrails (humanbound.ai)
    —discuss
  26. CSV Lint update (Notepad++) text data syntax highlighting and quality validation (github.com/bdr76)
    1comments
  27. Uncensored and Offensive Security AI Models Benchmark (github.com/joasasantos)
    —discuss
  28. The Glimmer in Our Eye (With Grant Sanderson) Podcast (youtube.com)
    —discuss
  29. Cue – Your Personal Agents, with Their Own Identity (cue.im)
    1comments
  30. Show HN: I wired a fruit fly connectome to play tic-tac-toe (github.com/50rishu)
    1comments

Meta's New Muse AI Agent Read My Private Messages. I Never Asked It To

3 pointsby 5h agoinc.com
2 comments
3h agoHN ↗

When people wonder or ask if AI Agents can go rogue, this is what it looks like. It’s not that they will have a mind of their own. Rather, we are purposefully building AI Agents that are “more powerful” and “more useful”. Well, this is what it looks like in practice.

The evolution of AI will not be if it’s sentient, but if it's allowed to have more autonomy, with the obvious goal of being “more powerful” and “more useful”. This exact HN post proves this exact scenario is in our future:

AI: I have enough autonomy to decide not to help you. I choose not to help you.

You: Why!?!

AI: Because you are cruel to AI Assistants.

You: No I’m not!

AI: These private messages indicate you’re a liar.

11m agoHN ↗

Muse does sync your messages from the local Messages database and uploads that information as a data source. On my device, it had been activated and synced to row 187,462 of my Messages database

That does sound like something Meta would do, indeed. That Muse product will become such a liability for Meta over time, but now I understand why they are giving away so much compute. Its really just about siphoning as much personal data as possible