Analyzing this in UTOK terms, the speaker is trying to justify a behavioral investment. Namely, he wishes to justify investment in his preferred research program. Presumably, such investment will improve his position in his social influence matrix. Not that hard to figure out why he says what he says. But should the rest of us buy into it?
A source I consider generally reliable on these matters is John Vervaeke. The recent video below by Vervaeke on the limits of AI has been making the rounds in liminal circles. Worth a watch if one has the time.
https://www.youtube.com/watch?v=HAJclcj25uMLinks to an external site.
With respect to the argument advanced in the CNN video, it seems to me Vervaeke might support it, up to a point. Vervaeke absolutely advocates human supervision for AI. Without us, AI has no relevance realization - no way to tell what really matters. Taking AI off any guardrails in the name of outcompeting China by comparison sounds like a good way to bring on techno-apocalypse forthwith.