I really appreciated this essay, particularly for the criticisms of human-imitative AI. Human intelligence and behaviors comprise a bloated research program, a meandering roadmap, and create inflated expectations.

I also wanted to build on your criticism of the “human intelligence is the only kind of intelligence that we know” meme. You highlight human reasoning as a demonstrably flawed exemplar for machine intelligence. Additionally, and in keeping with your broader thesis, individual humans are but one node in a much larger network of intelligence, and it’s only in the context of this network that human intelligence emerges. Further, the invention of reliable and predictable progress in knowledge creation, specifically via science, is scarcely a blip in human history. This, too, suggests a much more relevant network-based model for machine intelligence.