Innovation illusion
How “innovation” obscures the human costs driving technological change
When researchers uncovered last year that the start-up Clearview AI had created a face recognition system to search against photos on the Internet, some said it might “end privacy as we know it.” By making it possible to identify nearly everyone, whether online or off, the company threatened to render anonymity obsolete. This was a privacy “nightmare scenario.” In collecting biometric data from millions of unconsenting individuals, Clearview could facilitate unprecedented levels of surveillance, jeopardizing free speech and civil liberties in the process.
Many people already knew about face recognition, but this new tool seemed even more advanced, potentially even more threatening. Was it a breakthrough technology that pushed the technical bounds of image analysis? An impressive, albeit frightening, innovation reflecting the possibilities of investment in technological advancement?
No. Clearview exists because it broke the terms of service of multiple Internet giants. Clearview’s “innovation” was to blatantly disregard the rules.
What Clearview offers is neither clever nor unique, but a result of its willingness to risk breaking the law and antipathy towards the consequences of doing so. The company built a database of faces by scraping public photos from YouTube, Facebook, Twitter, and other platforms, against those companies’ policies. It also built a face recognition algorithm to compare faces to this massive database. Countless other companies have developed face recognition algorithms or web scraping tools. Platform companies’ policies against data scraping may be primarily about protecting their own intellectual property, but they can also protect the privacy of people who use their products. Clearview is notable only insofar as it ignores this.
Not all “advancements” are the natural product of scientific discoveries, technical breakthroughs, or novel ways of thinking. Sometimes, they’re just the result of obscuring and displacing the costs of technological change. Sometimes, their success derives from their ability to avoid regulation. And sometimes, they merely “invent” things that already exist, or invent solutions to problems that don’t.
There’s a fundamental myth in America — and particularly in Silicon Valley — that technical tools can fix our complex political, social, and economic problems. Evgeny Morozov dubs this fatal critical fallacy “technological solutionism.” Blind faith in technology confers a supreme importance on the notion of “innovation,” mandating its pursuit in the name of “progress.” Technology is portrayed not just as an objective, autonomous counterpart to flawed humans but also “indistinguishable from magic.” It is given an almost religious reverence or deference, with “innovation” cast as its lifeblood. Humans are the problem, and technology is the solution. Thus, anything that encourages technological development is not just good, but imperative.
This is, of course, a myth. The consequences of indulging this fantasy are threefold: an inclination to ignore the sometimes painful costs of new technology; disincentive to address underlying issues in society, preferring technological “fixes”; and the tendency to overvalue “innovation” to the point of seeing it where it doesn’t exist.
A complex set of political, social, and economic arrangements determine who gets to “innovate” and what. Even so, the “revolutionary,” “disruptive” tools we end up with are often just reconfigurations of rules and norms. One doesn’t have to look far to see this throughout the tech and tech-adjacent sectors. Uber and other ride hailing apps are essentially unregulated taxi services that can maintain such competitive rates and nimble flocks of drivers because they burn through VC money and don’t need to provide their “contract” workers with benefits. The cost is merely passed from users onto drivers (and VCs, for the time being). Airbnb, similarly, operates a de facto housing and hotel market, unencumbered by traditional taxes and regulations.
Yet, a collective illusion projects an inflated sense of ingenuity on the whole “innovation” enterprise. Technologies are marketed and celebrated as “AI,” autonomous, intelligent marvels of modern science, while the underpaid, invisible workers who train and power the underlying algorithms are erased. Meanwhile, many “AI” start-ups don’t even use “AI” in any meaningful way, appropriating the label to secure investors and users. When growth is the imperative, phantom “innovation” becomes ubiquitous. When every new product or service must be radical or game changing, the essence of what exactly is new about it and why is less important than the fact of its disruption.
Growth can be accompanied by growing pains: deskilling, unanticipated harmful uses, unemployment. Business leaders paint these consequences as unpleasant but necessary externalities that can be managed in order to ensure continued progress. But in the quixotic quest for growth, it’s easy to accept increasingly tenuous surrogates for progress. The negative side effects of Clearview’s product are clear: the empowerment of stalkers, increased authoritarian surveillance capabilities, the chilling of free speech associated with the loss of anonymity. But it’s also important to remember the product even exists in the first place simply because Clearview ignored regulations that protect people from Orwellian surveillance.
A society that overvalues technological “innovation” is bound to seek it at all costs, deploy it hastily on societal problems in lieu of structural reform, and find its expression in dubious places. Such an environment discourages critical analysis of not just the potential unexpected consequences of a new technology, but whether there’s a reason the technology hadn’t existed until that point in the first place. Often technological advancements result not from technical achievement but from managerial decisions that exploit regulatory gaps, obscure and displace costs, or repackage existing products as new and original. Yet these tactics are exempted from scrutiny under the umbrella of “innovation,” and any consideration of what has been flouted or disregarded becomes an afterthought. No technology has an absolute value. Innovation is not a good in and of itself. To properly evaluate the worth of any new technology, it’s essential to ask: is there a reason it didn’t exist until now? What costs are it displacing, and onto whom? What configurations is it replacing, and what new ones is it creating?
Jameson Spivack is an Associate with the Center. You can follow him on Twitter at @spivackjameson.