Gave this a brief skim, and it's paradoxically both highfalutin and often times plain wrong at the same time.
> Computers are universal machines. I use this term in a precise sense: anything that can be computed in the universe can in principle be computed by the kind of machine that we already have, given enough memory and time
Yeah, this precise sense is precisely wrong. There are plenty of things computers can't do (and will never be able to do), including (probably) be conscious (see Searle), but maybe even more difficult to swallow, generally compute entropy[1], or compute the spectral gap[2].