Turn: Antisocial innovation: most innovations produce no social value and some actively harm people- bump stocks, tobacco companies, pricing algorithms, surveillance technologies, and harmful AI prove.

According to all these perspectives, the biggest concern with innovation is that society might not get enough of it. Without massive government funding or monopolistic patent pricing, we may miss out on the next round of innovations that will increase economic growth, propel society forward, and improve human well- being. But what if that’s not the whole story? What if it turned out that a substantial portion of innovative activity was actually bad for society as a whole? In this Article, we make just this argument. While we are, of course, thrilled to be able to research and write this Article on internet-enabled computers, with access to clean water, indoor plumbing, antibiotics, vaccines, and a host of other modern conveniences, we don’t view these transformational innovations as typical. No doubt, many innovations have radically transformed human lives, prolonging and improving them. But we shouldn’t let these breakthroughs distort our view of innovation as a whole. Vanishingly few innovations dramatically improve society, while many don’t really affect it at all. And, we argue, a distressingly high number make life worse. We call these antisocial innovations.

Almost no innovations are purely antisocial, in the sense that they produce only harms and no benefits. Bump stocks, which enable semi-automatic firearms to operate like fully automatic weapons, may come close, in the sense that they have no meaningful military use and exist primarily to evade regulation of machine guns.25 More typical are innovations where the benefits to one group are outweighed by the costs to other interests. Consider the following examples.

When California prohibited cigarette manufacturers from adding flavorings, including menthol, to their products, some producers adopted a synthetic chemical that generates menthol’s cooling sensation but that does not add unlawful flavor.26 Obviously, this change is valuable for the cigarette manufacturers, but it is almost certainly going to reduce the effectiveness of the flavor ban and, accordingly, social welfare.27

Pricing algorithms are another example. For centuries, merchants relied on their own judgment to set prices for their goods.28 They were able to consider their costs, estimate their customers’ willingness to pay, and perhaps apply a limited knowledge of what their competitors were charging. Before the twentieth century, merchants typically did not advertise their prices; instead, prices were based on the outcome of bargaining with individual customers.29 With the advent in the twentieth century of convenient methods for displaying and changing prices, most retail prices were standardized: almost all consumers paid the same price for the same goods.30 Now, many sellers, including the most powerful merchants—Amazon, Walmart, and the big airlines—rely on computer programs to set their prices.31 These pricing algorithms generate prices based on traditional inputs like cost of goods and demand, but they can also account for the prices of multiple rivals, the time of day, day of the month, and time of year a sale is made, and, with the advent of big data, they can personalize prices for individual consumers.32 This technology is a huge step forward for sellers, but on the whole consumers will almost certainly pay more for algorithmically priced goods.33 The same is also true for apartment rents, which may be getting more expensive because large landlords use algorithms to set prices.34 A significant wealth transfer from consumers to sellers is the most likely result, reinforcing already alarming wealth distribution disparities in the United States.

Consider too the vast array of surveillance technologies that companies are employing in the workplace to track laborers’ motion, bathroom breaks, keystrokes, and eye movements.35 While firms may claim that surveilling workers will make them more efficient and, ultimately, reduce consumer prices, technology’s impact on workers’ health, privacy, and autonomy is alarming. Employees at Amazon warehouses face greater physical risks than do police officers, 36 and surveilling workers’ behavior is especially damaging for labor organizers, women, and people from historically marginalized communities.37

Finally, take artificial intelligence (AI). Recent stunning breakthroughs in AI have created tools, like ChatGPT, that can mimic human speech and writing or create accurate images based on text.38 AI is likely to have a range of positive impacts for society, including automating mundane tasks humans currently perform, creating new forms of art, and helping solve challenging technical problems in science and medicine. But AI has great potential to harm society. Its ability to take on many human tasks may lead to mass unemployment. And AI machines are likely to act in unpredictable ways, even in ways that physically harm humans. Those who know AI best are among the most concerned. A group of leading AI experts, from Google, Open AI, and other firms involved in developing the technology, signed a statement warning that “[m]itigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”39

As these examples attest, many innovations are, on balance, harmful to people. While we don’t deny the value of lifesaving breakthroughs like the Covid-19 vaccine, it is important for policymakers to recognize that many, and perhaps most, innovations do not fit the rosy image offered by politicians and tech gurus. This Article begins to challenge that image, and it offers a renewed vision of policymaking in a world where innovation can be beneficial but where it can just as easily be awful.