Some of the observations may seems a bit simplistic or overly general, as computer-science journalism often does.

But I like the counter-point between increasingly sophisticated pattern recognition and apparent stagnation in other AI fields/tasks and I think there’s a lot to think about here.