I've been reading his Understanding AI publication for a while. In my experience he's a good journalist in general, but has a massive blind spot when it comes to advanced AI, and frequently makes elementary errors in service of his already-decided conclusion to downplay them. I still think it's worth reading for the other stuff; his reporting on the state of self-driving cars has been quite good, for example.
https://tenor.com/en-GB/view/oh-no-anyway-gif-19022587