A.I. and technological determinism: Will A.I. exacerbate inequality or destroy it? And do we have a say?
In a recent New York Times op-ed, Elisabeth Mason, who runs Stanford’s Center on Poverty and Inequality, argues that Artificial Intelligence has the potential to be a powerful weapon in the war on poverty by allowing us to customize education, match candidates to opportunities, and predict future workforce demands. But, while it’s true that A.I. could indeed be used in this manner to combat unemployment and perhaps therefore poverty (although as Thomas Piketty and others have shown, it’s wealth, not income, that poses the greatest threat to equality), increasing integration of A.I.-enabled software and automated hardware will also enable technology to replace and obsolete both jobs that exist and, preemptively, those that are likely to.
Now most people would agree with Mason that it’s up to us to decide how we design and deploy artificially intelligent systems. We can use A.I. — and here I refer only to weak A.I., which focuses on a single task but can’t rival human intelligence — to elevate and empower the human workforce or we can use it to exploit and eliminate it. Underlying this outlook is the belief that technology is socially constructed, that is, a product of our cultural context. The technology we create is conditioned and driven by our values, biases, attitudes, and decisions in the aggregate. Although this may seem self-evident, there are some who would disagree.
Technological determinists, for instance, would contend that it is not principally up to us how technology evolves. For the determinist, the necessity of using technology to gain a competitive edge, whether it’s over another company or another country, will ensure that technology is used in an exploitative manner. Gunpowder is a prime example. Gunpowder wasn’t invented to subjugate other civilizations, but its invention allowed civilizations to do so. Thus, the determinist would say that it’s technology that shapes our culture and economy, not the other way around.
Karl Marx is often credited with fathering this view. “The handmill,” Marx wrote in The Poverty of Philosophy, “gives you society with the feudal lord: the steam-mill, society with the industrial capitalist.” Technology evolves, and as it evolves it drives social change. Yes, human beings are necessary to effect that change, but we are merely instruments acting in accordance with the technological and competitive forces that surround us.
If we accept this deterministic premise for the sake of argument, A.I. seems destined to exacerbate poverty and inequality in the long run. The prevailing technological paradigm is still, as it was on Marx’s time, predicated on the competitive quest for profit and productivity. Companies invent or adopt technologies in order to make money, using them to boost sales or cut costs, which often means laying people off. Technology offers many benefits, but those benefits have accrued (at least in the developed world) to a relatively small number of technologists and investors.
Indeed, automation has already contributed to the growing equality gap in the developed world by outmoding blue collar labor. Although it’s true that new jobs in technology have been created by these trends, automation is starting to replace those jobs as well. Imagine how much larger the inequality gap will be when A.I. obsoletes white collar jobs too — and, lest the creative class think it’s safe from rise of the machines, A.I. is already starting to produce groundbreaking works of art.
For Marx, technology was a weapon that opened up new opportunities for exploitation. Today, instead of the bourgeois and the proletariat, society is being quickly divided into what philosophers Alexander Bard and Jan Söderqvist called the netocracy, the new elite whose power is built on technology, and the consumtariat, the underclass who consume technology produced by the netocracy. In the determinist formulation, A.I. will create new ways for the netocracy to both obsolete workers and pacify consumers, thus exponentially intensifying inequality as we ‘Netflix and chill.’
If you think this deterministic view of A.I. seems pretty grim, I agree. I tend to believe, like Mason, that we have a say in what technologies humanity develops and how we use them. It’s true that the rise of netocracy and growing inequality fueled by technology are real and seemingly inexorable headwinds that are becoming more and more palpable every day. But I blame these issues on what Rensselaer Polytechnic professor Langdon Winner calls technological somnambulism, the notion that people approach technology passively and apathetically. This psychological distance, which Winner calls “sleepwalking,” prevents people from recognizing the larger implications of technology, giving rise to unmindful decisions and unintended consequences.
If Winner is right and this is the case, then before A.I. can truly be used to effectively address systemic issues like poverty and inequality, we need to get policymakers, corporations, and consumers to think more holistically and systematically about technology. We, as a society, need to consciously recognize and decide that the evolutionary trajectory of technology is not deterministic, but that it is within our control if we work together intentionally and mindfully.
Hopefully, as co-founder of Google-owned AI company DeepMind Mustafa Suleyman has said, 2018 will be the year a serious discussion on A.I. ethics and the ethics of A.I. (and there’s a distinction) finally takes place. Without it, we may continue to sleepwalk right into a technologically determined future.
About the author
Remington Tonar is a Partner and innovation consultant at Brandsinger, a NYC-based strategy consulting firm. He holds graduate degrees from NYU and Loyola University Chicago, and is currently writing his PhD dissertation on technological myth.
