One question we often hear in the tech industry is: Will A.I. destroy jobs and ultimately the society in which we live?
The fact that A.I. could one day become smarter than humans is not the main issue here. What people mean when they express this concern is: “Will bots replace us?”
In many countries, the service industry is responsible for most of the GDP, and people feel safe simply knowing that machines couldn’t do what they do — which has been the case for a long time. During the Industrial Revolution, machines did replace some workers in factories and on farms, but even now, people are needed to run and maintain the machines.
Yet there is a shift — bots are getting smarter.
Now, we see smart machines beating human players at Go. These machines are becoming more and more intelligent, which raises a lot of concerns. Indeed, the World Economic Forum released a report that describes what it calls the “fourth industrial revolution” and the loss of over 5 million jobs — a revolution that includes A.I.
So how should we really feel about this? Here are some useful suggestions.
A.I. will work with humans, not against them
A.I. won’t replace humans in jobs where responsible actions need to be taken. Our society is built around the fact that we know how to handle the human mistake but not the machine’s. We’ve seen this at Julie Desk, where I’m the founder. Clients are actually more demanding with Julie, who is A.I.-based, than with a real assistant! We admit that a real person has flaws, and we’re used to forgiving them. But with machines, we expect no errors.
And here’s an important point to understand: A.I. systems, compared to automations we use in factories, are built with statistics (to keep it simple) — which means that you cannot expect absolute perfection from A.I. Mistakes will happen eventually, even with the smartest of A.I. systems.
For Julie Desk, we try to remedy this by ensuring human supervision. This allows us to bring together the best of the two worlds, where a human still has the final say before an email is sent, for example. In a way, we see this as creating a new job title: A.I. trainer/supervisor. What the A.I. will do is help you make more accurate and efficient decisions.
Furthermore, we’ve actually been using AI for a long time — maybe without even realizing it. For example, in airplanes, A.I.-based autopilots stand in for the pilot. While A.I. helps improve the way pilots do their jobs, pilots and copilots ensure that the A.I. is doing its job correctly! And, in our society, we’re certainly not ready to accept an A.I. system crashing a plane while the pilot is away from the cockpit. The pilot would definitely get the blame for having left.
How to embrace disruptive innovation
A.I. is a disruptive technology, and disruptive technology creates a market. “Established companies are held blinded by their customers and ignore emerging markets of buyers,” explained Clayton Christensen in his book The Innovator’s Dilemma. The A.I. is actually doing something that nobody else has done before.
In our case, Julie Desk primarily offers a service for those who don’t already have an assistant. Those who do have assistants employ Julie to take care of the time-consuming, mechanical task of scheduling, allowing their real assistants to display other skills that benefit the company, grow the business, and can perhaps help them gain a promotion.
We believe that the fourth industrial revolution, with its disruptive innovations, requires that we prepare ourselves rather than panic. But reports like the one the from World Economic Forum don’t take that tack. We see more alarming articles and declarations about the danger of A.I. and new technologies than reflections on how to embrace this revolution and benefit from it.
Indeed, we need to keep an open mind: A.I. won’t so much destroy jobs as influence the “end of [the] human era.” We should use A.I. in a cautious and thoughtful way so that we don’t lose control over what we’re doing and why we do it.
A.I. should assist humans in what they do, without being an entirely autonomous system. (The only autonomous systems we know for now are actually more automated than intelligent — trains with no driver, for example.) If they did become entirely autonomous, though, we would probably find ourselves facing what we call an Ex Machina dilemma: Should A.I. be considered a new race we need to live with, as equals? It’s a debate that has a far larger dimension than the one related to losing jobs.
So let’s keep things simple and accept A.I. as a place where human and machine intelligence meet. In a way, just as eyeglasses are part of our individuality, just as a prosthesis helps a person become more mobile, A.I. can be seen as a way to boost individuals and help them be more autonomous and efficient. We should consider A.I. as a generous and positive evolution of the individual.