This is less about the article linked by Yann, but more about what Yann writes. He basically uses the argument made in the article, and adopts it to make a similar argument about AI. And that this argument about AI is the interesting content here (although the linked article is also interesting).