USA $12.00/CAN $13.00
N AT I O N A L A C A D E M Y O F S C I E N C E SN AT I O N A L A C A D E M Y O F E N G I N E E R I N G
I N S T I T U T E O F M E D I C I N ET H E U N I V E R S I T Y O F T E X A S AT D A L L A S
A R I Z O N A S TAT E U N I V E R S I T YS P R I N G 2 0 1 4
ISSUES
INSC
IENC
EA
ND
TECH
NO
LOG
YSPRIN
G 2014
VOLU
ME XXX N
UM
BER 3
IN SCIENCE AND TECHNOLOGY
90 ISSUES IN SCIENCE AND TECHNOLOGY
rently done by humans, includingknowledge work, and this transforma-tion will be on the scale of the first In-dustrial Revolution.
There are however, major flaws intheir framework. First, it’s not clear thatMoore’s law will continue to be true.Gordon Moore’s revolutionary predic-tion in 1965 that the number of tran-sistors on a chip would double every12 to 18 months (and thus would com-puter processing speeds) has provenprescient. Indeed, over the past 40years, processing speeds have increasedover 1 million-fold, unleashing a waveof innovation across industries. Butpossibly as soon as 2020, the dominantsilicon-based CMOS semiconductorarchitecture will probably hit physicallimits (particularly pertaining to heatdissipation) that threaten to compro-mise Moore’s law unless a leap can bemade to radically new chip architec-tures. But B&M devote no attention tothis critical issue, blithely assumingthat semiconductor past is prologue.
Second, after asserting that Moore’slaw will continue—not just in semi-conductors, but in all areas of digitaltechnology—they argue that we are ex-periencing “the digitization of just
in the entire world. There’s only oneproblem with their utopian claim: Toreach their projected income levels theworld would have to experience pro-ductivity growth of 25% a year for thenext 25 years, up from the 3.5% aver-age of the past 25 years.
Like Diamandis and Kotler, Bryn-jolsson and McAfee are similarlyutopian, arguing that the “second ma-chine age” (the first one was the Indus-trial Revolution in Great Britain) is“doing for mental power…what thesteam engine and its descendants didfor muscle power. They’re allowing usto blow past previous limitations andtaking us into new territory.”
Clearly one (or both) of these campsmust be wrong. We can’t be simulta-neously facing stagnation and surge.What this points to is the difficulty inaccurately describing the “innovationelephant.” Optimists see the parts thatare accelerating and driving change(e.g., our smart phones) and extendthem to the entire economy while ex-trapolating current trends forward.Pessimists see parts of the innovationsystem that are “stuck” (e.g., much ofthe personal and knowledge serviceseconomy) and assume that this notonly describes the entire economy butwill not change going forward. The re-ality is that neither view is right, be-cause some parts of the innovation sys-tem are driving rapid change, whereasothers are relatively stagnant.
Brynjolfsson and McAfee (B&M) inparticular assume that virtually all partsof the innovation system are vibrantand accelerating. They assert that in-novation will accelerate at an exponen-tial rate because of three factors: con-tinued exponential advances in com-puting power, pervasive digitization,and the combinatory nature of inno-vation. For them, these three factorsare enabling transformative tools thatwill replace large amounts of work cur-
Choosing a futureThe Second Machine Ageby Erik Brynjolfsson and AndrewMcAfee. New York, NY: W. W.Norton, 2014, 320 pp.
Robert D. Atkinson
The past several years have witnesseda lively debate about innovation be-tween techno-pessimists and techno-optimists. The pessimists’ view—ex-emplified by work such as Peter Thiel’sWhat Happened to the Future, RobertGordon’s The Demise of U.S. EconomicGrowth, and Tyler Cowen’s The GreatStagnation—is that the days of robustU.S. innovation and productivitygrowth are over, in part because mostof the low-hanging fruit has alreadybeen picked. Gordon, for example, as-serts that “there is no need to forecastany slowdown in the pace of future in-novation for this gloomy forecast tocome true, because that slowdown al-ready occurred four decades ago.” Forhim, “medical research, small robots, 3-D printing, big data, and driverless ve-hicles” are marginal extensions of pasttechnologies which will do little todrive future growth.
Confronting the innovation pes-simists are the innovation optimists,exemplified by, among others, PeterDiamandis and Steven Kotler’s Abun-dance: The Future is Better Than YouThink and Erik Brynjolfsson and An-drew McAfee’s The Second MachineAge: Work, Progress, and Prosperity ina Time of Brilliant Technologies. ForDiamandis and Kotler, emerging tech-nology is so powerful that “within ageneration, we will be able to providegoods and services, once reserved forthe wealthy few, to any and all whoneed them.” They don’t just meanabundance in the developed world, but
sciencepolicy.asu.edu • [email protected]
working on creative solutions to many of today’s most complex challenges: climate change • sustainabilityenergy • security • public healthinformation privacy • democracyglobal developmentinfrastructure
arizona state
university
one year master’s
science & technology policy
SPRING 2014 91
to better understand and improve ourworld.”
But this is a simplistic view of theprocess of innovation that likens it torandom recombinations of elements,akin to having a million monkeys ontypewriters, hoping that one will writea Shakespeare play. If this were true,the rate of innovation should have spedup over the past 100 years as morebuilding blocks were created and morepeople were at work combining ingre-dients. But innovation is no faster to-day than it was in the late 1800s. Infact, it appears that innovation is get-ting much harder, because the prob-lems to solve are so much tougher, andthe only thing keeping us from suffer-ing an innovation drought is the in-creased global resources going to R&D.
This leads them to perhaps theirlargest misreading of the future, onethat is shared by many futurists speak-ing at corporate confabs, TED-talk
coming at your expense. This counter-intuitive property, however, applies toprobably less than 5% of the economy,and certainly not to activities such asmaking cars or waiting on tables, wherescarcity and rivalry are the rule.
Third, B&M assert that innovationis speeding up because the possiblecombinations of innovations are in-creasing, as is our ability to combinethe ingredients. For them, innovationis easier in the digital era because ofthe possibility to recombine “recipes”and test them, what they call “recom-binant innovation.” They claim that the“number of potentially valuable build-ing blocks is exploding around theworld.” Growth is being held back onlyby our inability to process all the newideas fast enough. In short, they arguethat the “second machine age will becharacterized by countless instances ofmachine intelligence and billions of in-terconnected brains working together
about everything.” In other words, notonly are digital technologies improv-ing exponentially, but more and moreareas of the economy are becomingdigital. For them this matters because“when things are digitized…they ac-quire some weird and wonderful prop-erties. They’re subject to differenteconomies, where abundance is thenorm, rather than scarcity.”
Although it is true that digital tech-nologies are reshaping traditional in-dustries, including transportation,manufacturing, education, and healthcare, this does not mean that bits willreplace all atoms or genes. Food won’tbe digitized. Manufactured goods, al-though increasingly sold online andmade with digitally enabled technolo-gies, will still be made of atoms. WhatB&M are really referring to is digitizedinformation, where abundance is realbecause digital goods are nonrivalrous,meaning I can enjoy them without that
B O O K S
pundits, and pretty much everyonewho works at Silicon Valley’s Singular-ity University: the notion that techni-cal progress is improving “exponen-tially.” If innovation is actually improv-ing exponentially every few years, thiswould suggest that a decade from now,the U.S. Patent Office should be issuing4.4 million patents a year, up from the542,000 in 2013 (the exponential rateof increase). I can’t wait.
Finally, they overstate the extent towhich digital innovation is transform-ing occupations. For them, virtually alljobs will be disrupted by smart ma-chines. A closer look suggests other-wise. In a back-of-the-envelope analy-sis of U.S. occupations, the Informa-tion and Technology InnovationFoundation came up with a roughly 20-50-30 split among jobs that are mod-erately difficult to automate, difficultto automate, and very difficult to auto-mate. In other words, only about 20% oftotal U.S. jobs are likely to be easily au-tomated over the next decade or two.
Despite the utopian future that B&Msuggest is waiting for us, at least interms of innovation, they are surpris-ingly pessimistic about the effects,warning of all sorts of dystopian results,the principal one being massive unem-ployment and income inequality. Theyhave backed off somewhat from theirmore extreme claims made in their e-book The Race Against the Machine, inwhich they claimed that the second ma-chine age would cause massive unem-ployment. But they still raise the fearflag, arguing that “as computers getmore powerful, companies have lessneed for some kinds of workers.”
But their logic is fundamentallyflawed. For example, after pointing outthat since the year 2000, productivityand employment are no longer grow-ing apace, they assert that this disjunc-ture is evidence that productivity killsjobs. But not only has productivity not
accelerated after 2000, there is simplyno logical reason why growth in thelabor force and growth in productiv-ity would be related.
And although they acknowledgethat technologically driven productiv-ity would reduce prices, which shouldenable consumers to purchase more ofother goods and services, thereby em-ploying more workers, they dismissthis possibility. Without any evidenceor logic, they claim that consumers willbe satiated and not want to consumemore even if their disposable incomesgo up. I don’t know about MIT profes-sors, but the average U.S. family with ahousehold income of around $50,000would be ecstatic if higher productiv-ity doubled or even tripped their realincomes and would easily find thingsto spend it on.
The most disturbing aspect ofB&M’s argument is that it might leadpolicymakers to conclude that their jobshould be to slow down innovation-driven productivity growth. B&M ar-gue that “we can do more to inventtechnologies and business models thataugment and amplify the unique ca-pabilities of humans to create newsources of value, instead of automat-ing the ones that already exist.” Theyadvocate that government award prizesfor technologies that don’t replace la-bor. They want to start a “made by hu-mans” labeling movement. And sincetechnology will destroy jobs and cre-ate a massive new lumpenproletariatsitting at home with nothing to do, theyadvocate for a slew of redistribution-ist, rather than growth, solutions, in-cluding a negative income tax, an ex-panded Earned Income Tax Credit, anda national mutual fund that providesdividends for everyone.
The excesses of the techno-optimistsdo not mean, however, that the techno-pessimists are correct in asserting thatwe can no longer expect much bene-
fit from innovation. Given the slowgrowth in U.S. productivity over thepast decade and the large expected risein retirees over the next quarter cen-tury, the most important thing policy-makers can do is support innovationsthat “automate the jobs that already ex-ist.” Stoking neo-Luddite fears of tech-nology-induced joblessness is a step inthe wrong direction.
Robert D. Atkinson ([email protected])is president of the Information Technol-ogy and Innovation Foundation inWashington, DC.
Steal this bookOpen Accessby Peter Suber. Cambridge, MA: MITPress, 230 pp.
Paul F. Uhlir
In 1971, Abbie Hoffman mischievouslynamed his first book-length screedSteal This Book, and founded a pub-lishing company, Pirate Press, becauseno existing publisher would touch it.It was a countercultural manifestoagainst the “pig” establishment. Thenetworks—CBS, NBC, and ABC, plusthe upstart Fox—were “evil corporateconglomerates” spewing capitalist lies.
Flash forward four decades, and theinmates are running the asylums. Es-tablishment journals, including Issuesin Science and Technology, publish theircontent freely and openly online, invit-ing readers to “steal” their articles. Theold TV networks still exist, but the onethat really matters is the global Inter-net, where free information rules. Theborn-digital generation regularly piratescopyrighted content or at least expectsinformation to be free and instanta-neous. And there are even political “Pi-
92 ISSUES IN SCIENCE AND TECHNOLOGY
