AI self-improvement timeline moved up
The story so far
Anthropic founder Clark has moved the timeline for AI self-improvement (RSI) from 2028 to 2027. This is based on TasteVal benchmark test results, which measure the computational efficiency of AI models in scientific research tasks. Anthropic's Opus 5.5 model achieved an expert level with a 2.3 computational multiple. Since December 2025, the research taste of top AI models has doubled every three months. This change indicates that the potential of AI self-improvement is accelerating, although the specific realization still needs time to be verified.
AI-written from the reports · Updated 24 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- The DecoderAnthropic co-founder Jack Clark moves his prediction for self-improving AI forward to 2027
Anthropic founder Clark, based on TasteVal benchmark test results, has moved the timeline for AI self-improvement (RSI) from 2028 to 2027. TasteVal measures the computational efficiency of AI models in scientific research tasks. Anthropic's Opus 5.5 model achieves expert level with a 2.3 computational multiplier. Since December 2025, the research taste of top AI models has doubled every three months.