I'm not sure what issue others had with your comment. You're quite correct that he ignores vast swaths of current ML art and attacks a narrow conception of what ML is. Many of his criticism are legit with the right caveats, but he leaves out a lot of information that could lead to a different thesis.
My read of the discussion here is that there is lots of idle speculation by people who don't have any real experience with ML research / engineering, that overwhelms a minority who actually know what they are talking about and are calling CD out on this, or at least challenging aspects of his arguments.
My read of the discussion here is that there is lots of idle speculation by people who don't have any real experience with ML research / engineering, that overwhelms a minority who actually know what they are talking about and are calling CD out on this, or at least challenging aspects of his arguments.