Will big data algorithms soon control our lives?

By Stephan Manning.

Just a few decades ago it was unthinkable that a computer could ever be as smart as a human. But in 1996, the super-computer Deep Blue beat world champion Gary Kasparov in chess, and in early 2017 Google’s AlphaGo defeated the best human player in Go. In his fascinating new book Homo Deus, Yuval Harari argues that the combination of big data and self-improving algorithms will soon outsmart humans entirely and make human decision-making obsolete. Even today, it just takes 150 Facebook likes for psychometrics software such as Cambridge Analytica to know your needs, fears and hopes better than your parents do, and just over 300 likes for such software to know you better than you know yourself.  All based on analyzing your likes against Millions of other likes and profiles. No wonder the Trump campaign made effective use of that software last year to better target their voters. But this is just the beginning: Recently, researchers from Wits University in Johannesburg, South Africa, were able for the first time to directly link a human brain to the Internet – creating the first ever ‘Brainternet’. Based on increased connectivity, smart algorithms may soon be able to monitor and analyze all our biological functions, thoughts, interactions, and purchases, and know much better what we want and what makes us happy than we do. Harari argues that in the end humans may delegate all important decisions – choices of careers, partners and places to live – to algorithms that exceed our brain capacity manifold. So will big data algorithms eventually control our lives?

Every technology push in recent history has led to great hopes and fears. When humans invented railroads and telegraphs in the 19^th century, many rightly believed that accelerated transportation and long-distance communication would increase mobility, facilitate transactions and lead to economic prosperity. But some, like the famous author Henry Thoreau, would be much more critical. Will we lose touch with nature and our immediate surroundings? And what if “we are in great haste to construct a magnetic telegraph […], [but] have nothing important to communicate”? When computer numerical control (CNC) machine tools were widely introduced in factories in the 1970s, companies were looking to reduce costs, increase efficiency and improve quality control. But critics argued that such automation would lead to the ‘deskilling’ of machinists, and make work much less meaningful. Going forward, the economist Paul Krugman argues that the current wave of service automation may lead to an irreversible loss of jobs. And indeed we see it coming: “chatbots” replacing human call centers; robots running hotels; intelligent software composing classical music and writing the next Game of Thrones book. Science fiction authors would go even further and predict that computer technology might become self-aware and seize control over inferior humans entirely. Classics such as “2001: A Space Odyssey” and “The Terminator” all tell us similar stories of the future.

Harari’s argument follows the same pattern. While big data and algorithms will increase our cognitive capacity, help us make better choices, and prevent diseases and disasters, they might gradually take over our lives and replace ‘free will’ and ‘intuition’, since each individual will try to optimize their choices, minimize chances of failure, and maximize their potential.

But there is a different way of interpreting what’s going on. In his earlier bestselling book Sapiens, Harari argues that what makes us humans special is not our cognitive capacity, emotions, and creativity, not to even mention our very limited sensory and physical abilities. There is sufficient evidence of highly intelligent octopuses and whales; mammals and even plants having feelings; chimpanzees that can paint; countless species with superior vision, hearing, and smell that could easily kill or outrun us. No, what makes us humans special is our ability to imagine things, communicate in complex ways and coordinate ourselves in large numbers. No animal we know of has ever established a religion, imagined any utopia or apocalypse, or set up any sophisticated norms, laws or ethical codes of behavior that are followed by Thousands or Millions.

Now interestingly, most if not all advanced technologies humans ever invented helped us compensate for our many limitations compared to animals, but they never touched any of our true strengths. Trains made us faster; planes made us fly; machinery made us stronger; computers made us smarter; medicine made us live longer. And no doubt these accomplishments have made humans more productive and led to economic prosperity. But how about things that give us purpose and meaning? Was any religion ever ‘enhanced’ through technology? Did any significant value, human right or legal regulation result from smart algorithms? Did computers ever produce social movements or political systems?

Social innovations such as democracy or human rights, as much as social hierarchies and inequalities, are not just very complex in practice, but their development is typically very messy and nonlinear. Something that smart algorithms are neither able nor designed to predict. Throughout history, humans have started wars, destroyed the environment, and produced immense injustice. But humans have also made social progress – often in response to catastrophic events. For example, the American with Disabilities Act (ADA), which gives people with disabilities better access to jobs and public services, was passed by Congress in 1990 based on pressure from a social movement led by wounded veterans from the Vietnam War. The European Union, with all its flaws, emerged from the traumatic experience of World War II. Advanced technology had very little to do with those social innovations, nor was it responsible for the emergence of ISIS and Boko Haram. Surely, terrorism of today utilizes social media to spread fear and anxiety, but the roots and causes are much deeper.

So how much will big data algorithms change all that? I agree with Harari that they may have a huge impact on how individuals, corporations and governments make decisions. They might even affect the direction of political campaigns, social movements and terrorist groups. Algorithms may also influence increasingly which news people consume, especially on social media – even if human editors remain important gatekeepers of information. But will algorithms also affect the very social fabric of human societies? Or will rather ‘old-school’ structures, such as social hierarchies, economic inequalities, and political institutions prevail? Many argue that big data technology did not develop ‘automatically’ but has been driven by rather conventional economic interests of corporate Internet giants like Google and Facebook and security interests of nation-states, and that their use may reproduce and even reinforce already established hierarchies and inequalities. Harari himself gives a fine example. Who will benefit and who will suffer from accelerated technological upgrading of our human body, e.g. through nanobots regulating our health? Winners certainly include medical device producers and wealthy customers, whereas the poor may fall further behind, unless some law comes along to mitigate such inequalities. If so, who is to blame for such consequences? Smart algorithms, or the hierarchies of privilege and power that have co-evolved with human societies for thousands of years?

And how likely will big data algorithms actually replace our ‘free will’? Harari suggests that individual freedom in the future might also be the freedom to disconnect, the freedom to use smart technologies that suit our lives and avoid those that might harm us or get us addicted. Or as Henry Thoreau put it: “A man is rich in proportion to the number of things which he can afford to let alone.” But this requires education and may produce a new divide – between those conforming to big data algorithms to make decisions for them, and those who are able to lead more autonomous lives.

So maybe big data algorithms – and those who develop and commercialize them – will not control what drives and constrains us in life, but they may evoke new opportunities and social orders that influence our choices and challenge our individuality.