He's created a huge following (and is presumably making a lot of money) from pushing a hardcore AI-skeptic narrative, and I can't blame him for seeing that opportunity and running with it. We're ultimately all responsible for recognising these people and weighting their advise as necessary.
Additionally, from a public reputational perspective making bad predictions simply doesn't matter. In finance we're all aware of perma-bears who will predict the sky is about to fall, and when it doesn't just argue that the disaster is still coming, but is taking longer than expected, or that some unforeseeable thing happened which has compounded the risk but has for now kicked the can down the road.
So ultimately, it will be very hard to say Zitron is wrong unless he starts time-boxing his predictions, which I don't believe Zitron has done for obvious reasons.
That said I don't listen to him much at all. I've tried to listen to him a few times, but it's become evident very quickly that he doesn't understand the technical details enough to be making the predictions he's making and seems way too emotionally invested in the arguments he's pushing without good reason. I have strong personal filters for low-quality sources like Zitron – if someone raises enough flags I avoid them like the plague.