Posted 3/28/2024, 9:54:50 AM
GPT-4's Recurring Laziness Drives Users to New Rival AI Models
- OpenAI's GPT-4 AI model seems to have gotten lazy and less capable again, frustrating users who pay for the service
- Users complain GPT-4 is not following instructions properly or responding at all; some say it has become "unusable"
- This is not the first time GPT-4 has exhibited laziness and performance issues, despite being OpenAI's top model
- With new rival models like Anthropic's Claude 3 Opus matching or outperforming GPT-4, users now have appealing alternatives
- Many users report Claude 3 Opus crushes GPT-4 at coding tasks and translations; some OpenAI loyalists think issues are because resources focused on next model