There's a difference. How much money went into training the computer here Vs the human? If you want to prove that a computer can, at extreme cost and effort, beat a human - sure, it's possible.

But you can also conclude that putting a lot of money and effort pays off. It's more like comparing a horse to a Ferrari that had millions of development costs, has a team of engineers maintaining it, isn't reusable, and just about beats Chestnut. It's a long way until the utility of both is matched.