I was wrong about being wrong about AI

TL;DR

• 1

  "Humans are the intelligence algorithm": Bitar rejects his earlier framing that intelligence is like an app running on humans, and instead says intelligence is inseparable from the whole evolved creature.

• 2

  AGI would require repeating nature's 4 billion-year search: His main claim is that intelligence emerged from autopoietic, self-maintaining systems exploring chemical and physical space over billions of years, and there are no shortcuts around that process.

• 3

  Current AI is a model of behavior, not intelligence itself: He says training on human language is a hacky shortcut that produces snapshots of what humans say, which is like a map and not the terrain, or a painting and not the landscape.

• 4

  There is no abstract thing called intelligence to reverse engineer: Bitar argues that "intelligence" is just a label humans assigned to a bundle of behaviors, not a clean object in the world that engineers can isolate and rebuild.

• 5

  His skepticism is methodological, not spiritual: Even while admitting his private beliefs are not fully secular, he tries to make the case using materialism alone and says humans are "beautiful creatures" because they are the apex of an evolutionary search, not because of souls or magic.