This is a frustrating book, because I simultaneously agree with the conclusion and an entirely unimpressed with its argument. There is some great stuff here about the meaning of "reality" (things that are _real_ are those things that are involved in our best explanation for a phenomena), and it gives a good heuristic metric for the quality of arguments (good arguments are ones in which there are few-to-none variables which can be changed without affecting the outcome.) Also that people are relevant in the grand scheme of the universe, due to their engineering and scientific prowess to effect change. I liked all that stuff.
What I didn't like was Deutsch's bad logic. He plays fast-and-loose with shaky metaphors, and generalizes their arguments back to the real world. For example, he takes issue with Haldane's quote that "the Universe is not only queerer than we suppose, but queerer than we can suppose." I've always interpreted this as either "the universe is infinite, and we are not, therefore we simply can't enumerate all thoughts about its queerness," or as "something about our neural architecture makes us fundamentally blind to some of the universe's mysteries." Analogously to the first case, it's uncomputable to choose an arbitrary real number --- there's simply too many of them to ever possibly find most of them.
But this is all stupid, says Deutsch! He dismisses it metacircularly, saying that if the universe did have physically-possible but technically-impossible aspects, then "this fact would itself be a testable regularly in nature. But all regularities in nature have explanations, so the explanation of that regularity would itself be a law of nature, or a consequence of one. And so, again, everything that is not forbidden by nature is achievable, given the right knowledge." As is so often the case with physicists, philosophers, and traditional mathematicians, Deutsch is missing the capital-C Complexity issue of this problem. Just because something must have an answer doesn't mean it must be computable, which is to say, doesn't mean it is necessarily instantiable inside the universe. As an illustration, there is an answer to "what is the quantum state of the universe," but the answer to that is too big to fit into the universe, other than _as the universe qua itself._ More generally, if a problem space grows too quickly, you simply can't expect to find an answer.
Most of the book is Deutsch having interesting ideas with extremely weak justifications. It gives off the impression that he wants to talk about his (in my opinion, not-so) crazy beliefs, but feels that he ought to justify them first. But either he couldn't be fucked to give good justifications, or his entire thought process is needlessly muddy and if he's gotten to the right answers, its purely by coincidence. As someone who already agreed with his conclusion, I found myself put off by his arguments; they simply would not compel a disbeliever to start believing, nor do I think they hold much water even stripped of the rhetoric.
All in all, I wanted to like this book, but put it down around 30% because Deutsch didn't seem to have a point, and his arguments were frustrating enough that I wasn't convinced I could trust him of anything novel he might have to show me tangentially.