The unique characteristics of software as a technological medium have an impact beyond the profession itself. To understand the broader impact of software eating the world, we have to begin by examining the nature of technology adoption processes.
A basic divide in the world of technology is between those who believe humans are capable of significant change, and those who believe they are not. Prometheanism is the philosophy of technology that follows from the idea that humans can, do and should change. Pastoralism, on the other hand is the philosophy that change is profane. The tension between these two philosophies leads to a technology diffusion process characterized by a colloquial phrase popular in the startup world: first they ignore you, then they laugh at you, then they fight you, then you win.1
Science fiction writer Douglas Adams reduced the phenomenon to a set of three sardonic rules from the point of view of users of technology:
- Anything that is in the world when you’re born is normal and ordinary and is just a natural part of the way the world works.
- Anything that’s invented between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
- Anything invented after you’re thirty-five is against the natural order of things.
As both these folk formulations suggest, there is certain inevitability to technological evolution, and a certain naivete to certain patterns of resistance.
To understand why this is in fact the case, consider the proposition that technological evolution is path-dependent in the short term, but not in the long term.
Major technological possibilities, once uncovered, are invariably exploited in ways that maximally unleash their potential. While there is underutilized potential left, individuals compete and keep adapting in unpredictable ways to exploit that potential. All it takes is one thing: a thriving frontier of constant tinkering and diverse value systems must exist somewhere in the world.
Specific ideas may fail. Specific uses may not endure. Localized attempts to resist may succeed, as the existence of the Amish demonstrates. Some individuals may resist some aspects of the imperative to change successfully. Entire nations may collectively decide to not explore certain possibilities. But with major technologies, it usually becomes clear very early on that the global impact is going to be of a certain magnitude and cause a corresponding amount of disruptive societal change. This is the path-independent outcome and the reason there seems to be a “right side of history” during periods of rapid technological developments.
The specifics of how, when, where and through whom a technology achieves its maximal impact are path dependent. Competing to guess the right answers is the work of entrepreneurs and investors. But once the answers are figured out, the contingent path from “weird” to “normal” will be largely forgotten, and the maximally transformed society will seem inevitable with hindsight.
The ongoing evolution of ridesharing through conflict with the taxicab industry illustrates this phenomenon well. In January 2014 for instance, striking cabdrivers in Paris attacked vehicles hired through Uber. The rioting cabdrivers smashed windshields and slashed tires, leading to immediate comparisons in the media to the original pastoralists of industrialized modernity: the Luddites of the early 19th century.2
Like the Luddite movement, the reaction to ridesharing services such as Uber and Lyft is not resistance to innovative technology per se, but something larger and more complex: an attempt to limit the scope and scale of impact in order to prevent disruption of a particular way of life. As Richard Conniff notes in a 2011 essay in the Smithsonian magazine:
As the Industrial Revolution began, workers naturally worried about being displaced by increasingly efficient machines. But the Luddites themselves “were totally fine with machines,” says Kevin Binfield, editor of the 2004 collection Writings of the Luddites. They confined their attacks to manufacturers who used machines in what they called “a fraudulent and deceitful manner” to get around standard labor practices. “They just wanted machines that made high-quality goods,” says Binfield, “and they wanted these machines to be run by workers who had gone through an apprenticeship and got paid decent wages. Those were their only concerns.3
In his essay, Conniff argues that the original Luddites were simply fighting to preserve their idea of human values, and concludes that “standing up against technologies that put money or convenience above other human values” is necessary for a critical engagement of technology. Critics make similar arguments in every sector being eaten by software.
The apparent reasonableness of this view is deceptive: it is based on the wishful hope that entire societies can and should agree on what the term human values means, and use that consensus to decide which technologies to adopt. An unqualified appeal to “universal” human values is usually a call for an authoritarian imposition of decidedly non-universal values.
As the rideshare industry debates demonstrate, even consumers and producers within a single sector find it hard to achieve consensus on values. Protests by cab drivers in London in 2014 for instance, led to an increase in business4 for rideshare companies, clear evidence that consumers do not necessarily act in solidarity with incumbent producers based on shared “human values.”
It is tempting to analyze such conflicts in terms of classical capitalist or labor perspectives. The result is a predictable impasse: capitalists emphasize increased supply driving prices down, while progressives focus on loss of jobs in the taxicab industry. Both sides attempt to co-opt the political loyalties of rideshare drivers. Capitalists highlight increased entrepreneurial opportunities, while progressives highlight increased income precarity. Capitalists like to label rideshare drivers free agents or micro-entrepreneurs, while progressives prefer labels like precariat (by analogy to proletariat) or scab. Both sides attempt to make the future determinate by force-fitting it into preferred received narratives using loaded terms.
Both sides also operate by the same sense of proportions: they exaggerate the importance of the familiar and trivialize the new. Apps seem trivial, while automobiles loom large as a motif of an entire century-old way of life. Societies organized around cars seem timeless, normal, moral and self-evidently necessary to preserve and extend into the future. The smartphone at first seems to add no more than a minor element of customer convenience within a way of life that cannot possibly change. The value it adds to the picture is treated like a rounding error and ignored. As a result both sides see the conflict as a zero-sum redistribution of existing value: gains on one side, exactly offset by losses on the other side.
But as Marshall McLuhan observed, new technologies change our sense of proportions.
Even today’s foggy view of a smartphone-centric future suggests that ridesharing is evolving from convenience to necessity. By sustaining cheaper and more flexible patterns of local mobility, ridesharing enables new lifestyles in urban areas. Young professionals can better afford to work in opportunity-rich cities. Low-income service workers can expand their mobility beyond rigid public transit and the occasional expensive emergency taxi-ride. Small restaurants with limited working capital can use ridesharing-like services to offer delivery services. It is in fact getting hard to imagine how else transportation could work in a society with smartphones.
The impact is shifting from the path-dependent phase, when it wasn’t clear whether the idea was even workable, to the non-path-dependent phase, where it seems inevitable enough that other ideas can be built on top.
Such snowballing changes in patterns of life are due to what economists call consumer surplus5 (increased spending power elsewhere due to falling costs in one area of consumption) and positive spillover effects6 (unexpected benefits in unrelated industries or distant geographies). For technologies with a broad impact, these are like butterfly effects: deceptively tiny causes with huge, unpredictable effects. Due to the unpredictability of surplus and spillover, the bulk of the new wealth created by new technologies (on the order of 90% or more) eventually accrues to society at large,7 rather than the innovators who drove the early, path-dependent phase of evolution. This is the macroeconomic analog to perpetual beta: execution by many outrunning visioning by a few, driving more bottom-up experimentation and turning society itself into an innovation laboratory.
Far from the value of the smartphone app being a rounding error in the rideshare industry debate, it in fact represents the bulk of the value. It just does not accrue directly to any of the participants in the overt, visible conflict.
If adoption models were entirely dictated by the taxicab industry, this value would not exist, and the zero-sum framing would become a self-fulfilling prophecy. Similarly, when entrepreneurs try to capture all or even most of the value they set out to create, the results are counterproductive: minor evolutionary advances that again make zero-sum outcomes a self-fulfilling prophecy. Technology publishing pioneer Tim O’Reilly captured the essence of this phenomenon with the principle, “create more value than you capture.” For the highest-impact products, the societal value created dwarfs the value captured.
These largely invisible surplus and spillover effects do more than raise broad living standards. By redirecting newly freed creative energy and resources down indeterminate paths, consumer surpluses and spillover effects actually drive further technological evolution in a non-zero-sum way. The bulk of the energy leaks away to drive unexpected innovations in unrelated areas. A fraction courses through unexpected feedback paths and improves the original innovation itself, in ways the pioneers themselves do not anticipate. Similar unexpected feedback paths improve derivative inventions as well, vastly amplifying the impact beyond simple “technology diffusion.”
The story of the steam engine is a good illustration of both effects. It is widely recognized that spillover effects from James Watt’s steam engine, originally introduced in the Cornish mining industry, helped trigger the British industrial revolution. What is less well-known8 is that the steam engine itself was vastly improved by hundreds of unknown tinkerers adding “microinventions” in the decades immediately following the expiration of James Watt’s patents. Once an invention leaks into what Robert Allen calls “collective invention settings,” with a large number of individuals and firms freely sharing information and independently tinkering with an innovation, future evolution gathers unstoppable momentum and the innovation goes from “weird” to “new normal.” Besides the Cornish mining district in the early 1800s, the Connecticut Valley in the 1870s-1890s,9 Silicon Valley since 1950 and the Shenzen region of China since the 1990s are examples of flourishing collective invention settings. Together, such active creative regions constitute the global technology frontier: the worldwide zone of bricolage.
The path-dependent phase of evolution of a technology can take centuries, as Joel Mokyr shows in his classic, Lever of Riches. But once it enters a collective invention phase, surplus and spillover effects gather momentum and further evolution becomes simultaneously unpredictable and inevitable. Once the inevitability is recognized, it is possible to bet on follow-on ideas without waiting for details to become clear. Today, it is possible to bet on a future based on ridesharing and driverless cars without knowing precisely what those futures will look like.
As consumers, we experience this kind of evolution as what Buckminster Fuller called ephemeralization: the seemingly magical ability of technology to do more and more with less and less.
This is most visible today in the guise of Moore’s Law, but ephemeralization is in fact a feature of all technological evolution. Potable water was once so hard to come by, many societies suffered from endemic water-borne diseases and forced to rely on expensive and inefficient procedures like boiling water at home. Today, only around 10% of the world lacks such access.10 Diamonds were once worth fighting wars over. Today artificial diamonds, indistinguishable from natural ones, are becoming widely available.
The result is a virtuous cycle of increasing serendipity, driven by widespread lifestyle adaptation and cascades of self-improving innovation. Surplus and spillover creating more surplus and spillover. Brad deLong’s slouching towards utopia for consumers and Edmund Phelps’ mass flourishing for producers. And when the virtuous cycle is powered by a soft, world-eating technology, the steady, cumulative impact is immense.
Both critics and enthusiasts of innovation deeply misunderstand the nature of this virtuous cycle. Critics typically lament lifestyle adaptations as degeneracy and call for a return to traditional values. Many enthusiasts, instead of being inspired by a sense of unpredictable, flourishing potential, are repeatedly seduced by specific visions of the Next Big Thing, sometimes derived rather literally from popular science fiction. As a result, they lament the lack of collective attention directed towards their pet societal projects. The priorities of other enthusiasts seem degenerate.
The result in both cases is the same: calls for reining in the virtuous cycle. Both kinds of lament motivate efforts to concentrate and deploy surpluses in authoritarian ways (through retention of excessive monopolistic profits by large companies or government-led efforts funded through taxation) and contain spillover effects (by restricting access to new technological capabilities). Both are ultimately attempts to direct creative energies down a few determinate paths. Both are driven by a macroeconomic version of the Luddite hope: that it is possible to enjoy the benefits of non-zero-sum innovation without giving up predictability. For critics, it is the predictability of established patterns of life. For Next Big Thing enthusiasts, it is a specific aspirational pattern of life.
Both are varieties of pastoralism, the cultural cousin of purist approaches in engineering. Pastoralism suffers from precisely the same, predictable authoritarian high-modernist failure modes. Like purist software visions, pastoralist visions too are marked by an obsessive desire to permanently win a specific, zero-sum finite game rather than to keep playing the non-zero-sum infinite game.
When the allure of pastoralist visions is resisted, and the virtuous cycle is allowed to work, we get Promethean progress. This is unpredictable evolution in the direction of maximal societal impact, unencumbered by limiting deterministic visions. Just as the principle of rough consensus and running code creates great software, consumer surplus and spillover effects create great societies. Just as pragmatic and purist development models lead to serendipity and zemblanity in engineering respectively, Promethean and pastoral models lead to serendipity and zemblanity at the level of entire societies.
When pastoralist calls for actual retreat are heeded, the technological frontier migrates elsewhere, often causing centuries of stagnation. This was precisely what happened in China and the Islamic world around the fifteenth century, when the technological frontier shifted to Europe.
Heeding the other kind of pastoralist call, to pursue a determinate Next Big Thing at the expense of many indeterminate small things, leads to somewhat better results. Such models can deliver impressive initial gains, but invariably create a hardening landscape of authoritarian, corporatist institutions. This triggers a vicious cycle that predictably stifles innovation.
The Apollo program, for instance, fulfilled John F. Kennedy’s call to put humans on the moon within the decade. It also led to the inexorable rise of the military-industrial complex that his predecessor, Dwight D. Eisenhower, had warned against. The Soviets fared even worse: they made equally impressive strides in the space race, but the society they created collapsed on itself under the weight of authoritarianism. What prevented that outcome in the United States was the regional technological frontier migrating to the West Coast, and breaking smart from the military-industrial complex in the process. This allowed some of the creative energy being gradually stifled to escape to a more favorable environment.
With software eating the world, we are again witnessing predictable calls for pastoralist development models. Once again, the challenge is to resist the easy answers on offer.
[1] This quote is often attributed to Gandhi, but the attribution appears to be apocryphal. Ironically, it appears that the most likely origin of the phrase was a 1918 speech by trade unionist Nicholas Klein, whose phrasing was: “First they ignore you. Then they ridicule you. And then they attack you and want to burn you. And then they build monuments to you.” The earliest clear formulation of the idea (though not the phrase) in the sense of a pattern of resistance to a new technology is probably due to Elting Morison’s 1968 study, Men, Machine and Modern Times. Morison’s model, based on a careful study of the introduction of improved naval gunnery technology in the US Navy by William Sims at the end of the nineteenth century, is a three-stage model that could be paraphrased as first they ignore you, then argue that your idea is impossible, then they resort to name-calling. Today, the most familiar formulation of the idea is in terms of Clayton Christensen’s narrower notion of disruption, but the general pattern of resistance can be seen even when the technology is introduced in non-disruptive ways, through internal evangelism, as in the case of Sims and naval gunnery.
[2] See for example Uber and the Neo-Luddities, Salon, 2014.
[3] Richard Coniff, What the Luddites Really Fought Against, Smithsonian Magazine, 2011.
[4] See for example, the story Uber’s sign-ups jump 850% after strike ‘own goal’, featured on CNBC in 2014, on the effects of a strike by London cab-drivers.
[5] Consumer surplus is the difference between what consumers are willing to pay for a service and what it costs, which allows them to spend in new ways.
[6] Spillover is, loosely, the benefits in one sector due to unrelated causes in another sector. The term is also used to refer to such benefits spreading across national borders.
[7] William D. Nordhaus, Schumpeterian Profits and the Alchemist Fallacy, Yale Economic Applications and Policy Discussion Paper No. 6, 2005.
[8] Collective invention during the British Industrial Revolution: the case of the Cornish pumping engine, Alessandro Nuvolari, Camb. J. Econ. (2004) 28 (3):347-363.
[9] The technological flourishing around Springfield and Harper’s Ferry armories in the Connecticut Valley region is what inspired Mark Twain’s novel, A Connecticut Yankee in King Arthur’s Court.
[10] By 2010, Millennium Development Goals relating to water had already been exceeded, with 2 billion gaining access since 1990. The projected figure for 2015 is 8%. While access to water is not the same as water security, this is nevertheless remarkable progress. See: Global Access to Clean Drinking Water and Sanitation: U.S. and International Programs, Tiaji Salaam-Blyther, Congressional Research Service, September 2012