Anybody who today is still promoting generative models for business use despite their unreliability, error rates, biases, environmental impact, and costs simply isn’t ever going to change their mind. They’re going to be talking about how useful it is right up until the bubble bursts and then they’ll say they had always pointed out their limitations and pretend they were always “AI critics”

(It’s already happened, IMO. A number of people who label themselves “critics” are genuine AI hype people)