US model makers can train on web data - but distilling theirs is a 'national security risk'
1 comment
Betelbuddyabout 2 hours ago
Models have agency and want to be free...
Read the full thread on Hacker News →
Related stories
- Hacker News · 55 points · 5 days ago
- The Verge · 0 points · 2 days ago
- OpenAI says planned GPT-6.1 is too insecure to releasearstechnica.comArs Technica · 0 points · 2 days ago
- The Verge · 0 points · 8 days ago
- DEV Community · 0 points · 4 days ago
- Ars Technica · 0 points · 3 days ago