Is AI-powered Automation Really to Be Feared, or Is It Our Best Hope?
Thoughts about digital transformation for legal & compliance advisors
These posts represent my personal views or occasionally those of colleagues working on legal and regulatory compliance issues related to enterprise digital transformation powered by the cloud and artificial intelligence. Unless otherwise indicated, they do not represent the official views of Microsoft.
Last week the cable TV channel HBO released a new documentary film, The Truth About Killer Robots. I confess I haven’t had time to watch it yet. But I have watched the trailer (click on the link above) and read favorable reviews in prominent publications like the UK’s Guardian. So I’ll be sure to watch it soon.
The provocative title and some equally provocative statements by the director, Maxim Pozdorovkin, make me wonder about the film’s underlying assumptions. These statements and the reviews make clear that the film is not really about “killer robots.” That’s only a hook to draw attention. The film does recount some well-known incidents where machines have killed humans—two Tesla accidents that happened while the cars were on autopilot, and a factory robot at a VW plant in Germany that killed an assembly worker.
But, according to the director, the real subject of the film is automation. Pozdorovkin believes that automation powered by robots and AI will lead not simply to loss of jobs, but to “de-skilling” and “the loss of human dignity associated with traditional labor.” This is a common idea these days, and many even regard it as inevitable. It is this sense of inevitability that I want to question.
The future of employment and human dignity in a world where intelligent automation is spreading into every organization and business process is an essential subject for debate. But the truth is that no one can predict the future with certainty. We can only try to forecast the future by careful study of the present and the past. The spread of AI-powered automation is not something that can be prevented, because the economic gains to society are too important. Yet AI is sure to drive disruption of existing business models that will be painful for many whose livelihoods are tied to those models. But if we learn the right lessons from past episodes of disruption driven by automation, and if we have the foresight to reinvest some of the gains from AI into smoothing the transition for those most at risk, we can end up in a better place.
My key point is that automation is not new. The disruption to existing ways of living and working caused by new technologies has been a recurring feature of human society for thousands of years.
In a broad sense, disruptive technology is an ancient and indissociable part of human nature. We humans have been a tool-making species from the beginning. The oldest known stone tools date to more than 2.5 million years ago. That’s more than 2 million years before the emergence of modern homo sapiens somewhere between 100,000 and 300,000 years ago. It is well established scientific fact that the use of tools—and more broadly of technology, which includes things like the controlled use of fire for tasks like cooking—has been a fundamental driver of human evolution. I believe that AI, far from being a radical break with the past, is only the latest incarnation of humanity’s urge to use tools and technology to build a better life for itself.
Stone hand axes and cooking fires are to be sure very primitive forms of automation. But the development of more sophisticated forms of automation—again, I’m using the word in the broad sense of “technology applied by humans to do useful work”—is also an ancient phenomenon. The first draft animals were domesticated around 3,500 BCE, immediately revolutionizing agricultural productivity and, according to scholars, creating considerable social upheaval as they did so. And agriculture wasn’t the only “business process” to be affected. Soon after horses were domesticated they were carrying warriors into battle on wheeled chariots. Kingdoms that possessed this terrifying new war machine overwhelmed their foot-bound enemies and established vast new empires such as those of Egypt and the ancient Hittites.
In 19th century America horse-based mechanization of work was a fundamental component of both rural and urban economies. In a fascinating blog post entitled The Day the Horse Lost Its Job, Microsoft’s President Brad Smith (a man of many talents) recounts how:
“It’s difficult to overestimate the degree to which the American economy and broader society revolved around horses. As one historian has commented, ‘every family in the United States in 1870 was directly or indirectly dependent on the horse.’ In rural areas, farmers prospered in no small measure by growing hay to feed the nation’s 8.6 million horses, or one horse for every five people.”
Of course, the horse economy of the 19th century was swept away by the automobile economy of the 20th. The rise of the automobile not only destroyed the livelihoods of all those whose jobs revolved around the horse, but even contributed to a brief but sharp depression in the early 1920s triggered by the impact on farm incomes of the collapse in demand for horse feed.
We know the automobile ultimately created many more jobs than it destroyed. According to McKinsey research cited by Smith, from 1910 to 1950 the rise of the auto destroyed 623,000 old jobs but created 7.5 million new ones.
This pattern of initial loss of jobs followed by much greater job creation has characterized the cycles of automation throughout human history. If history is a guide, it is reasonable to conclude that this pattern will continue to hold in the future as we witness the rise of AI-powered automation of all kinds.
Despite the political turmoil that surrounds us and seems to be a constant in human society, the world today is vastly wealthier than the world that saw the decline and fall of the horse economy just over a century ago. Not only in America, but in every region of the world, incomes are much higher and lives are much longer. Surprising as it may seem, statistics also show that we humans are getting less violent. Another fascinating blog post I came across recently, this one by Microsoft founder Bill Gates, publishes a chart from Harvard psychologist Steven Pinker showing that between “the late Middle Ages and the 20th century, European countries saw a tenfold-to-fiftyfold decline in their rates of homicide”:
Returning to our theme of economic growth powered by technological innovation, it is not pollyannish to point out that, as World Bank statistics document, a smaller percentage of humanity is living in extreme poverty today than at any previous time in history. At 10% that percentage is still too high. But I am persuaded that AI-powered automation and the tremendous new productivity it permits will be a key to reducing that number to zero one day in the not-too-distant future.
We are entering a time of uncertainty and change where we need to make sure that every member of society has the skills and resources to succeed. AI will play a crucial role in producing those resources. Our challenge is to use the lessons of the past to ensure that the next phase of humanity’s ancient and never-ending pursuit of automation benefits everyone.
Microsoft has published a book about how to manage the thorny cybersecurity, privacy, and regulatory compliance issues that can arise in cloud-based Digital Transformation—including a section on artificial intelligence and machine learning. The book explains key topics in clear language and is full of actionable advice for enterprise leaders. Click here to download a copy. Kindle version available as well here.