I can grant that there are people overhyping what today's generative AI can do, but this is very, very far from the worst case of overselling I've seen.
However, above and beyond what the existing stuff can do, I am also confident that it is technically-possible to build much more capable stuff, human-level AI. Doing that takes capital, so I am not shedding tears about investments being made towards AI R&D, just maybe the specifics of what individual efforts might be getting targeted.