I think the way he sketches the the AI labs as "marketing geniuses" for not just releasing their models as auto-correct is a bit cynical, as well as implying in general that these labs are muddying the waters on purpose by not agreeing with <authors position> and by engaging in "hype" (believing in the technology).