Two explanations for variation in human abilities

The chess analogy doesn’t work because those other experts not only spent similar amounts of hours on chess as Carlson but they were most likely people who were doing very well compared to others who played as much as them previously. Carlson’s performance is excellent compared to talented experts, not just to people who spent requisite amounts of time on chess. So the performance difference is much higher than what the author assumes.

Comparing the number of games played by AlphaGo vs those played by Lee Sedol also doesn’t work. AlphaGo was programmed to learn based on many games because a computer can process that many games and it works. But it doesn’t mean that it’s impossible for a computer to use different algorithms to extract information based on fewer games. Also, Lee Sedol has deeply studied many other games and their critical moves, not just the ones he played.