- He argued that it is better for AI researchers to focus on endowing computers with intelligence rather than working on areas like robotics or rule-based systems.
- Minsky suggested that "consciousness" does not really exist as a unit and that "consciousness" is simply a collection of many different processes in the brain that can eventually be figured out. I do not agree with Minskey on this point.
- He also suggested that genetic algorithms are mediocre and that the process of evolution is not smart by nature. When discussing the circuitry and organization in the brain, Minsky said that evolution will stoop to any level to get things done.
- Minsky also argued that the concept of "interior grounding" (which suggests that knowledge can be innately gained) is superior to the concept of "grounding in experience" (knowledge gained through contact with the outer world). He suggested that "interior grounding" can explain the innate Chomsky grammars in humans.
Minsky's talk was very interesting, and I'm glad that he came to UCI.
One can view the complexity of the brain (and the mystery of consciousness) through the evolutionary lens and through the lens of intelligent design. In my opinion, the design lens seems to give a more coherent and satisfying picture.