PERSON DIRECTORY
Keith
Keith appears in 2 indexed conversations across Machine Learning Street Talk. This directory brings every appearance, source, TL;DR, digest, and transcript into one searchable feed.
Karl Friston - Why Intelligence Can't Get Too Large (Goldilocks principle)
Friston’s Goldilocks principle places intelligence between randomness at tiny scales and complexity-erasing averaging at planetary or astronomical scales, while larger organizations require recursive structure and operate “on the edge of chaos.”Machine consciousness is possible in principle but may require embodiment, counterfactual breadth, temporal depth and substrate-dependent computation, directing attention toward processing-in-memory, memristors and neuromorphic photonics while leaving structural learning and causal-organization selection unresolved.
The Fractured Entangled Representation Hypothesis (Kenneth Stanley, Akarsh Kumar)
Tim ScarfeDr. DuggarKenneth O. StanleyAkarsh KumarKeith
PicBreeder shows that identical outputs can conceal radically different machinery: evolved networks factor a skull into reusable components, while conventional SGD produces “total spaghetti.”If fractured representations make adaptation, continual learning and transformative creativity costly, growing sparse, protected modules through open-ended selection could be dramatically more efficient, though the “trillion-dollar question” remains an unproven research agenda alongside continued LLM scaling.