That doesn't seem sensible. We already have general intelligences. Let's infer possible outcomes before rushing head first.
Also. That second point is like the most unhelpful point I've ever seen saying that "we just need to look at us, we're the real GI, we're proof AGI can exist". What are you even talking about? You don't think people have taken that philosophy before? We're not even close to figuring out all the nuts and bolts that go into /natural/ general intelligence. What makes you think it's easier here?
Nobody predicted LLM's? Despite the fact that it shares mechanisms from other older branches of machine learning. Right. I'd recommend you stop right here lol. You are no longer speaking from the point of view of the industry or research, just looking at what laymen think.