'We messed up': OpenAI panics over GPT-4o's overly obedient behavior - Ynetnews
... warns of real mental health risks —raising questions about whether its push for 'helpful' AI went too far.
Read the original at ynetnews.com ↗
- Topics
- Healthcare
- Mentions
- OpenAIGPT
- Tracked via
- “AI Risks”