2) No, actually? It's not all that "clear". The current net-beneficial impacts are highly contained to software development, and even there it seems to be of dubious financial value.
This is the whole problem. AI has to be even more worldchanging, even more used, even more endlessly profitable than the smartphone. Companies are massively overinvested now, if AI turns out anything less than that, they're all fucked. The need to make "the next iPhone" has meant no product is allowed to be merely "very good".
Culturally, AI has already lost. People have always snarked about the iPhone being a luxury product that's too expensive compared to the competition, but the fundamental premise was always accepted. Even in today's era of widespread recognition that people are "on their phones too much", the smartphone itself isn't up for debate.
They loathe AI. Palantir is singing the praises of AI as a tool for fascism. There have been multiple attempts to attack Altman's house because of the "AI is so impactful it'll kill us all!" rhetoric. People are furious about what just the hype around AI alone is doing to the job market.
And in the background, everyone is aware that it's a bubble. That even if AI did everything promised and more, that it is physically impossible to build enough datacenters quickly enough and that the mass-unemployment will annihilate the economy.