CARACAS — Artificial Intelligence is opening up new opportunities for the economy and society. But it will also affect millions of human jobs, and thus poses a huge challenge for public policymakers, warns a 2016 White House report on AI's projected impact on the U.S. economy.
Among other things, the report predicts that AI will have a positive impact on productivity growth in firms while dramatically shifting the skills people must have to participate in the job market. Job markets, the study argues, are expected to become fluid and uncertain as some positions disappear, giving way to others.
All of that makes foresight in public policymaking vital to making good use of AI's advantages. Policymakers must get ahead of the game, and design plans and compensations needed to mitigate the impact of the significant losses technological changes will entail for pertinent actors.
The report predicts that AI will have a positive impact on productivity growth.
As the projected changes mentioned above illustrate, AI and the economy are deeply intertwined. But to understand another dimension of the relationship between the two disciplines, one should consider the basic concepts of intelligence and AI.
The definition of intelligence is complex. Indeed, there are as many definitions as there are intelligent people. Just to keep things simple, let's consider intelligence as a wide-ranging set of mental abilities and potentials that become manifest in our actions.
AI also has many definitions, though it may be reduced here to the rational and logical action of machines in response to surrounding stimuli. Implied in that, of course, is an ability to learn constantly. AI only considers the rational (logical and mathematical) part of human intelligence, at least in its initial steps.
One of the pioneers of AI was Herbert Simon, a Nobel Prize winner in economics. Simon poked a number of holes in the principle of perfect rationality in economic agents, a fundamental premise of classical economic theory. First off, he argued that Homo economicus does not face situations in a state of perfect competence. Simon also showed how environmental uncertainty (imperfect information) prevents businesspeople from making exact predictions, forcing them to constantly revise objectives. This provokes costs that prevent them from reaching their optimal goal.
American political scientist, economist, sociologist, psychologist, and computer scientist Herbert Simon — Photo: TheFamousPeople
In addition, Simon notes that people are innately limited in their ability to perceive all of reality's complex elements. This "bounded rationality," as he calls it, is what justifies the use of machines to help us access all data needed to make optimal decisions. Rational machines can process vast amounts of data, far more than individual or even a group of people can, and can thus improve productivity and reduce costs.
For Simon, individuals are simple agents, and the complexity of their conduct has more to do with the complexity of their surroundings than their limited mental capacities. So if a person's rationality is limited and an individual cannot find optimal solutions to economic problems, firms will inevitably need machines to process vast amounts of data and find those solutions. In practice, this means procedural rationality, which is the sum of the abilities of many individuals in certain intelligent algorithms. In addition, AI forms are able to learn, be free of emotions and boost economic efficiency.
Simon was a pioneer of AI and its relationship with economics, in large part because he demonstrated the importance of the former by highlighting the later's limitations, and by critiquing the "perfectly rational" economic man. We know today that data is the driving force behind the Fourth Industrial Revolution, and that for firms to be competitive and productive, they need to properly process that data.
We also know that labor markets will very soon be determined by robotics, AI and all the disruptive technologies that empower production while substituting humans. This is an enormous challenge both for individuals and public policymaking within societies that must invest to maximize their capabilities and ensure they retain their roles in the face of competition from intelligent machines.