‘The Second Machine Age’

Reviewed:¹ Erik Brynjolfsson & Andrew McAfee, The Second Machine Age: Work, Progress and Prosperity in a Time of Brilliant Technologies, W.W. Norton & Co., New York, 2014. Hardcover, 320 pp.

“The factory of the future will only have two employees, a man, and a dog. The man will be there to feed the dog. The dog will be there to keep the man from touching the equipment.” —Warren G. Bennis²

“The hand-mill gives you society with the feudal lord; the steam-mill society with the industrial capitalist.” —Karl Marx³

In their new book, Erik Brynjolfsson and Andrew McAfee draw from the Industrial Revolution, its intellectual legacy, and recent research to set expectations for a new, digital revolution. Only now hitting full steam, this Second Machine Age will lead to a fresh transformation of the relations of production, with economic and social challenges no less drastic or far-reaching.

Their treatment is in three parts: the drivers of the digital revolution; its individual and economic consequences; and recommendations for the men and dogs looking to retain their jobs in Warren Bennis' factory (less caricature, it seems, than inevitability).

Two main drivers are exponential increases in computing power, represented by Moore's famous law, and the advent of big data—digitized information about the real world, with additional, wholly-digital layers. Crucially, information technology (IT) is identified as “general-purpose;” with improvements in business models, organizations and worker skills, it enables “recombinant innovation.” Brynjolfsson and McAfee present nuanced thinking on these factors, and argue persuasively that we should expect only shattered expectations and predictions falsified by astounding technical progress.

Next, through an accessible, math-free survey of labour, productivity and growth economics, the authors give implications of technological change for growth (labeled “The Bounty”) and inequality (“The Spread”). The Spread encompasses two dynamics. First, the winner-take-all nature of highly-scalable digital markets leaves second-best musicians, newspapers and social networks far behind number-ones, giving us society with not the industrial capitalist—but the global superstar. Second, the automation of a broad variety of routine jobs, physical or mental, endangers the 40-hour workweek of today's middle class.

So what's a poor, replaceable worker to do? Hope is found in a chess story: after Garry Kasparov was definitively bested by IBM's Deep Blue, ‘freestyle’ tournaments revealed that amateur players making savvy use of multiple machines could do even better. We must, then, learn to “race with computers,” instead of letting technology race ahead. The complementarity of human ingenuity and digital technology, in well-designed symbiosis, will unlock the full, recombinant potential of the Second Machine Age. What's more, labor scarcity will lead to decent wages for those physical jobs that resist automation.

Surely, this is all possible and desirable. But while the authors weakly disclaim technological determinism (p.255) and dismantle several happy illusions advanced by others,⁴ they fail to let on that the way is exceedingly narrow. Substantial blind spots hide threats on all sides.

These become clear in the third part, which lays out prescriptions for action. In order to race with computers, individuals are counselled to get the best education available, developing their uniquely-human creativity. Policy recommendations are, disappointingly, straight from Silicon Valley's latest lobbying playbook. U.S. immigration reform appears—H1-B visas for digital-economy workers but not, perhaps, Latin American laborers. The country's uneven public education system should be rebooted (with technology) to teach creativity—an admission that “best available” education is scarce. A discussion of tax reform dances around the question of whether the lopsided gains awarded by winner-take-all markets should be spread across The Spread.

The U.S. framing omits larger challenges. For instance, the authors correctly argue (Chapter 8) that the metric of GDP fails to capture a substantial amount of intangible value linked to digital interactions. Only Internet advertising has proved a viable way to monetize this activity; the king of online advertising is, of course, Google. Can enough new markets, and business models, be found for each country to have its Google, Baidu or Publicis? It's heartening to be told that intangibles have real value, but uncaptured value does not buy food.

The authors also observe that automation, by eroding the attraction of cheap labour, can “bring manufacturing back to America” (p.185). This is presented as benign. Will foreign workers be content to see winners in the United States, or China, take all? Will other nations—lacking robust, universal education that fosters creativity—greet “skill-biased technical change” warmly? In the global view, The Bounty predicted by Brynjolfsson & McAfee seems much smaller, and The Spread dauntingly wide.

The Second Machine Age will be most useful to those in digital business: corporate decision-makers facing the unenviable challenge of keeping their firms and (one hopes) employees afloat in a greatly uncertain future. Other readers can expect clear thinking on digital innovation and insights into the changing nature of work—but an incomplete, though skilful, elaboration of the far-reaching impacts. Brynjolfsson & McAfee have no solutions equal to the challenge…but then, who does?

(754 words, 77 in quotes)