All of which makes the book a delight: It's a high-wire performance, a feat of intellectual daring. He goes to war with almost everybody else who thinks about the Internet's people, institutions, and technologies in the public eye: Nicholas Carr, Clay Shirky, David Weinberger, Tim Wu, Kevin Kelly, Farhad Manjoo, Steven Johnson, Gary Wolf, among others. Sometimes, he wins easily. Sometimes, he tangles himself into knots trying to defeat every possible enemy and defend against every possible counterargument. In all cases, he is worth reading, even if you vehemently disagree.
The only comparable experience I've ever had was reading Vaclav Smil on energy: Frustrating, enlightening, and counterintuitive in the best meaning of the word. Of course, if you're interested in the Internet, you should read it. And I think historians in 2030 or 2050 will use this book to highlight the anxieties and debates of our time, pretty much all of which it attempts to tackle at once.
On the other hand, I'm not sure that they'll read it as a fair or grounded representation of the state of technology. Morozov's mode, while learned and theoretically grounded, is not as deeply authoritative as it appears. There is little actual evidence that many of the phenomena he highlights are actually occurring in the way he says they are. Granted, that's not his focus. I recognize fully that Morozov's project is in the realm of ideas, ideology, and the sociology of knowledge.
But there's often not even an attempt to line up reality with his anecdotes and projections. He relies time and again on scenarios, little flights of fancy, that are neither thought experiment nor forecast, but something more opaque. Even within the logic of the book itself, it's difficult to compare his scenarios to one another. There's little consistency among them in terms of plausibility or time-scale. If you look closely, you're left wondering: is this something that is already happening, might happen in a year, could happen in 10 years, or is a logical possibility in a century? Quite-close-to-real scenarios are delivered with the same rhetorical weight as truly wild Morozovian nightmares.
Let me give just a few examples.
Here's one section from Morozov's chapter on predictive policing in which he introduces a real product, ShotSpotter, a microphone sensor system that lets police (in Oakland, say) identify where gunshots are fired. Watch how he slides from there to a much stranger idea without blinking:
These systems are not cheap--ShotSpotter reportedly charges $40,000 to $60,000 a year per square mile--but they are hardly the latest word in crime detection. Why bother with expensive microphones if smartphones can do the job just fine? It all boils down to designing an appealing and nonintrusive app and creating the right incentives--perhaps by appealing to the moral conscience of citizens or by turning crime reports into a game--so that citizens can take on some of the tasks of faulty sensors and easily distracted human.
From an actual, deployed system to a nightmarish sousveillance scenario in one sentence. Could such a system work? Would this be appealing to institutional players or people? Why even bother, if you're the cops? Is anyone even thinking about doing this in one, five, or even 20 years? What gives him the idea this might happen? I don't know. There's certainly nothing in the quite extensive (and welcome) footnotes to explain this leap.