3 ms·
Arguably we're not that far from jumping up one level, having an AI agent that when encountering any new type of hard problem would train a new dedicated sub-AI
by falcor84 1y ago
Arguably we're not that far from jumping up one level, having an AI agent that when encountering any new type of hard problem would train a new dedicated sub-AI to solve that.
- moffkalast 1y agoDeep learning scientists automating themselves? Ironic.