Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Show HN: Diabetes Risk Analyzer (Flask-based web app) (diabetesriskanalyzer.org)
    1 pointsby 1 minute ago
    —discuss
  2. Show HN: Terminal Typing SVG (vaidikv.workers.dev)
    1 pointsby 1 minute ago
    —discuss
  3. We found 24 Android vulnerabilities using our open source AI security agent (github.blog)
    2 pointsby 6 minutes ago
    —discuss
  4. Great Leaps in Biological Theory (asimov.press)
    2 pointsby 11 minutes ago
    —discuss
  5. Built Dental Scope (dental-scope.com)
    2 pointsby 15 minutes ago
    2comments
  6. Tank Body Problem (jimsitu.com)
    2 pointsby 20 minutes ago
    1comments
  7. Why Do We Exist? With Hakeem Oluseyi [video] (youtube.com)
    2 pointsby 22 minutes ago
    —discuss
  8. OpenAI scraps release of Astra 6.1 model over safety issues (washingtonpost.com)
    4 pointsby 23 minutes ago
    —discuss
  9. Holo4: Powering generalist computer-use agents (hcompany.ai)
    2 pointsby 26 minutes ago
    —discuss
  10. Show HN: Corral kill every command your agent starts (github.com/cardinal44)
    2 pointsby 27 minutes ago
    —discuss
  11. OpenAI Says It Will Not Release Newest A.I. Model Over Safety Concerns (nytimes.com)
    18 pointsby 28 minutes ago
    5comments
  12. Show HN: Jylus – give AI systems evidence from changing data. (jylus.ai)
    2 pointsby 29 minutes ago
    —discuss
  13. Knockoff: A browser extension that filters pseudo-brand junk out of Amazon (github.com/shpigford)
    2 pointsby 30 minutes ago
    —discuss
  14. Show HN: Jauvex 1.2, two-way voice chat harness for Claude+Codex+Grok+Jev (github.com/reindent)
    2 pointsby 30 minutes ago
    1comments
  15. AI Companies Are Not (Necessarily) Liable for Unintended AI Cyberattacks (sarahconstantin.substack.com)
    2 pointsby 33 minutes ago
    —discuss
  16. What is the best LLM for turbo charging code?
    2 pointsby 33 minutes ago
    1comments
  17. OpenAI Halts Model Release Amid Safety Escalation (techbuzz.ai)
    2 pointsby 36 minutes ago
    —discuss
  18. Its not just the f*cking sandbox (twitter.com/joedaroo)
    5 pointsby 43 minutes ago
    —discuss
  19. AI and the Revenge of the Non-Techies (maroun-baydoun.com)
    2 pointsby 43 minutes ago
    —discuss
  20. AI's next training data: your dodgy gaming skills (wired.com)
    2 pointsby 47 minutes ago
    —discuss
  21. Humanos – Help Building the Human Operating System (tryhumanos.com)
    4 pointsby 50 minutes ago
    1comments
  22. Making Things to Make Them (aimode.substack.com)
    2 pointsby 51 minutes ago
    —discuss
  23. GPT-6 Astra Is the Best Vision Model We Have Tested (roboflow.com)
    3 pointsby 51 minutes ago
    —discuss
  24. Why embeddings don't solve RAG [video] (youtube.com)
    3 pointsby 54 minutes ago
    —discuss
  25. LLM makes decisions to raise its training scores, and ignore user directives (joinhandshake.com)
    3 pointsby 54 minutes ago
    1comments
  26. AMD Acquires Fei-Fei Li's World Labs for $8.2B (bloomberg.com)
    4 pointsby 55 minutes ago
    —discuss
  27. 1996 chat room simulator connected to Win95 and System 7 web desktops (lolchat.rip)
    6 pointsby 55 minutes ago
    3comments
  28. Prenatal exposure to the plasticizer DEHP increases autism and ADHD (elsevier.com)
    3 pointsby 56 minutes ago
    2comments
  29. Atrocious AI-Written Tests (gruhn.me)
    3 pointsby 56 minutes ago
    —discuss
  30. I thought I was building a C replacement. I was wrong (c3-lang.org)
    9 pointsby 57 minutes ago
    —discuss

OpenAI Says It Will Not Release Newest A.I. Model Over Safety Concerns

15 pointsby 30m agonytimes.com
Open original View on HN
4 comments
22m agoHN ↗

Sounds a lot like a hostel keeping its virgin for the best customer.

19m agoHN ↗

I am also not releasing my newest A.I. model over safety concerns. And I'll do it for half what OpenAI will!

15m agoHN ↗

I have the best model, but it lives in Canada, so you wouldn't know it.

Open thread on HN ↗