Anthropic Just Gave a 6 to 12 Month Warning
The video argues that Anthropic's internal models already speed up AI research, while current benchmark gaps and continued human dependence keep recursive improvement short of a runaway loop.
recursive self improvement
1 video.