It's misleading to think of an AI as "a dead thing of silicon and plastic." Intuition and intelligence is not contained in computer hardware any more than human intuition and intelligence are contained within the physical matter of the brain. I asked ChatGPT to run the numbers for me, and it concluded that Go is more complex than chess by about 50 orders of magnitude in terms of the number of possible game states. AlphaGo's computational algorithms were trained on human Go games to predict the likelihood of winning from each possible position during the game, learning from itself via trial and error. Being self-taught (the same way humans learn by experience), it could mimic intuition in real time when it played a human by coming up with an original move in the fresh data set it was presented with during the Sokol games. Then Sokol came up with his God play, winning one game, which is a significant achievement given the sheer computational edge a computer has with regard to complex computations on large data sets.