What strategy should one use to beat a deep learning machine in chess?

Tactically, there is probably very little a human being can do to defeat a chess-playing computer or a deep learning machine. The question asks for a strategy, which requires a different way of thinking.
Here is a suggestion. Since the computer chess programs are developed to defeat an opponent based on chess rules, and human opponents are inherently inferior because we do not have the computer’s computing prowess, the human strategy is to first acknowledge that fact, and stop trying to win against the computer opponent. Rather, the human strategy would be to play for draw every time. Devote all your energy to learn how to draw. Let the computer use up all its time to figure out an attack in a blocked position and then block it up some more. When the computer has less time to compute, maybe it won’t be able to make the best move. In any case, pat yourself on the back when you achieve a draw.
Another reason that humans cannot beat computers in chess is psychology. I have studied a few Grand-Master vs Computer matches, and you can feel that palpable fear in the human. The computer just marches on inexorably until the human breaks down and makes a stupid mistake. It’s hard to notice brilliance from the computer even when it wins. Playing for draw takes that pressure away for the human.
Since the computer plays to win while the human plays to draw, it may befuddle the computer. The deep learning computer may learn about this draw strategy and play for draw as well, hence human and machine will co-exist harmoniously. When humans discard their zero sum destroy your enemy mindset, we will not have to fear AI. In tennis terminology, it’s called Endless Love.