▲0Anthropic warns that it’s very close to RSI capability, aka self-building AI2 hours ago·0 comments·flag