So basically his book is a 304 page argument in favor of the lump of labor fallacy.
45 million Americans will lose their jobs to automation in the coming decades. Yep, sure. And those jobs "won't be replaced", so the obvious answer is to fund a UBI... somehow. Well, sorry to say, but rich folks are pretty good at hiding their money from the government, so they aren't gonna be paying for it. You won't risk a capital wealth tax because it will smother investments, and you won't get it out of businesses because they're taxed on net. No, middle class people would pay for it, and they're already too financially stressed to pay for extant services. (Insert your country here, because it's true across the whole world.)
You live in the Bay Area. It's a ****hole. Every unpainted house, every cracked, ****-stained sidewalk is work. There is no shortage of work, not even remotely, anywhere in the United States. There has probably never been so much work that each person could be doing.
The problem here isn't the "45 million Americans will lose their jobs" part, and it's not the "automation" part. It's the "those jobs won't be replaced" part. Because history has shown us countless times that rentiers have no incentive to invest, and Americans have been systematically denied access to those hoarded means of production, and have no way to create their own employment. That is the problem. Those jobs won't be replaced because capital has stopped allocating labor.
So even if you somehow solved all of the problems of a UBI - like the funding, like the inflationary effects, like the fact that slumlords are gonna triple their rent the second they know their tenants are making some money - it won't solve the real problem.