Leading AI Labs Say Autonomous Self-Improvement Is Near
Top AI labs are approaching recursive self-improvement (RSI), where models create the next generation. Anthropic said Claude performs 26% of research on its models, OpenAI plans an automated 'researcher' by March 2028, and Musk expects full automation at xAI by 2027.
- Claude conducts 26% of Anthropic's research on new models
- OpenAI promises automated AI 'researcher' by March 2028
- Musk: at xAI each model is created by the previous one, full automation by 2027
- OpenAI admitted it doesn't know how to safely achieve full RSI
Read next
AI