Baum on Age of Em

In the Journal Futures, Seth Baum gives the first academic review of Age of Em. First, some words of praise:

The book is by far the most detailed description of the em world available. .. If you are wondering about what some aspect of an em world might look like, odds are good that a description can be found somewhere in the book. .. The breadth of research covered is impressive. .. The Age of Em is a thoughtful and original work of futures studies, bringing an insightful social science perspective to the topic of mind uploading. The book covers wide ground about the nature of the em world, offering a valuable resource for anyone interested in the topic, or about futures studies more generally. The book is accessibly written and could be read by undergraduates or even advanced high school students, though experts from any discipline will also find much of interest. The book is especially worthwhile for anyone who could benefit from an overview of contemporary social science research, which the book provides in abundance.

I am pleased to hear this, and largely agree. But of course Baum also has criticism.

The book’s methodology is rooted in extrapolation of current social science to em world conditions. At times, the extrapolations seem strained. For example, a claim that economic benefits of urban agglomeration would continue in an em world (p.215) cites Morgan (2014), which is a popular media description of Amazon’s large data centers. It is true that Amazon data centers are large, and it may well be true that em cities are large for similar reasons, but the latter does not necessarily follow from the former.

I’m surprised Baum has any doubts that economies of agglomeration would continue. I’ve taught urban economics, and it seems to me that we understand in some detail many of the forces that push for and against the clumping of economic activities. Since one of the main forces resisting concentration, travel congestion, greatly reduces in an em world, while most forces pushing concentration remain strong, it seems to me a quite safe prediction that ems would clump together in cities.

In other stretches, Hanson’s personal tastes are apparent. This is seen, for example, in discussions of combinatorial auctions and prediction markets (p.184-188), two schemes for adapting market mechanisms for social decision making. Prediction markets in particular are a longstanding interest of Hanson’s. The book’s discussion of these topics says little about the em world and seems mainly oriented towards promoting them for society today. The reader gets the impression that Hanson wishes society today was more economically efficient and rational in a certain sense, and that he has embedded his hopes for a better world into his vision of ems.

It is a common habit of futurists to populate their imagined futures with many visions of more efficient physical technologies, but to presume no gains in social technologies. I think we should instead also expect the adoption of more efficient social technologies, especially in a more competitive world such as the em world would be. Which is why I tried to outline some changes of this sort. But it seems Baum prefers the usual habit.

The book proposes that those humans who own parts of the em economy “could retain substantial wealth”, but everyone else is “likely to starve” (p.336). To my eyes, this seems overly optimistic: if ems are so much smarter than humans, and if they have such a profit motive, then surely they could figure out how to trick, threaten, or otherwise entice humans into giving up their wealth. Humans would likely die out or, at best, scrape by in whatever meager existence the ems leave them with.

In history, there have always been some people who were much smarter than others, and yet the less smart have usually retained wealth, often great wealth. This remains true today. The correlation between individual wealth and intelligence is weak, and mostly not due to tricks or threats. Retirees today continue to control great wealth even though they tend to be physically, mentally, and socially less powerful. Clearly there isn’t a strong historical rule that the smartest take all wealth from the rest. Humans should probably be more concerned that the age of em would only last a year or two, and that we don’t know what happens next.

Given these dire prospects for humans, one might question whether it would be good to create ems in the first place. Unfortunately, the book does not consider this topic in any detail. The book is pro-em, even proposing to “subsidize the development of related technologies to speed the arrival of this transition [to the em era], and to subsidize its smoothness, equality, or transparency to reduce disruptions and inequalities in that transition” (p.375). But this position is tenuous at best and quite possibly dangerous.

I’m quite confident that the topic of evaluation of the em world will not be neglected, as so many seem so eager to discuss it. So I tried not to take much of a position in the book on the overall value of the em world, and a section toward the end of the book neutrally reviews many approaches to such evaluation. But it seems that readers will always try to find such a position in a book like mine, and then complain that the position is insufficiently defended.