Thinking Feed
Feed
Areas
AdminAgentic SoftwareAI Safety
You
Economics of recursive improvement: a threshold, not a prophecy

Backlinks

AI Safety — programmeupdated 26 OctTasteVal: a bounded human-beating experiment-design result, not autonomous research agenda settingupdated 10 OctCan measured research assistance compound into self-sustaining AI improvement?updated 4 OctAI-assisted AI research versus a self-sustaining recursive loopupdated 30 Sep