You have conflated job loss with superhuman intelligence. It doesn’t take a superhuman intelligence to vacuum my floor or move goods around a distribution center. Job loss has already happened, and it’s going to keep happening.
In contradistinction to this orthodoxy, I find the following five heresies to have more evidence to support them.
Intelligence is not a single dimension, so “smarter than humans” is a meaningless concept.
Your entire argument for this is an exercise in semantics. Would you say we are “smarter than ants?” Sure, ants can do some tasks better than we can, but it is intellectually dishonest to take the stance that we aren’t smarter than ants. ‘Smart,’ in this case, has a tacit, consensus meaning that encapsulates the mental capacity of various entities in a way that does not need precision and weighting on every potential axis of intelligence.
Humans do not have general purpose minds, and neither will AIs.
Emulation of human thinking in other media will be constrained by cost.
Dimensions of intelligence are not infinite.
Intelligences are only one factor in progress.
You are correct about all of this, but you have failed to demonstrate their relevance to your central argument. For instance, you cannot empirically predict costs and say whether those costs will prohibit superhuman AI. You can, at best, propose a testable hypothesis. Likewise, we know that dimensions of intelligence are not infinite, but you don’t have to make it to infinity to beat a 200 IQ.
I’m not saying superhuman AI will or will not happen; I am just saying that your arguments are just as speculative and unscientific as arguments that propose the inevitability of superhuman AI. I’m skeptical of all arguments about this that extend beyond about one iPhone cycle.
Side note, I don’t think you can say with certitude that spacetime is finite.