The last industrial revolution

 We are entering the last industrial revolution.

 Automatons – robots or programs will be impossible to for a human to outperform. It could be a good thing if we acknowledge this right now. But even if we don't, it will happen.

When I say "the last" I mean it in a metaphoric almost philosophical way. There may be further transformations in industrial processes, but the impact on humanity as a work force will be minimal. From a productivity point of view humans are about to be cut out of the picture. Further industrial transformations won't affect them.

Here I use the term “industrial” in its most basic definition: related to actual production. The production phase is when all design, QA testing and compliance with laws and regulations is done. Analyzing the need, designing a solution and making sure it is legal is part of the creation process. In the industrial phase of production, nothing should be surprising or require adaptability. The jobs bound to this phase are purely technical. And these purely technical jobs will disappear. With time it will be impossible for humans to compete against automatons. whether they are programs, A.I. or robots.

In fact, even the jobs that are not purely technical per se often carry with them a huge technical aspect. And automation will take over this aspect. People tend to think that automation will first affect simple, low wage jobs, and will creep up to more and more complex roles. That is not what is going to happen. For example there is little benefit in creating a robot to sweep the streets. And the task is a lot more complex than it may seem. A street is a constantly moving, constantly changing environment. Anything can happen in a street. It is much easier to create a robot for a controlled environment, with very limited potential for unexpected events. An operating room for example. Creating a system that can remove a brain tumor from an anesthetized patient is simpler than making a road legal autonomous vehicle that can clean the street, avoid road works, adapt its behavior to surrounding pedestrians and cars, and so on.

Are we there yet?

As of today only a competent MD can actually create a plan of action for the removal of a brain tumor. But how long before a robot can execute this plan in a way that is both faster and safer than anything a human could pull out. Does 15 years seems like a plausible answer? Because if it does, there is a problem.

It takes 12 to 15 years to become a licenced surgeon. So if 15 years is plausible, it means that some of the most brilliant and dedicated persons may be studying for naught. Starting on a path that will turn into a dead end right when they are done walking it.

I do not know about you, but I feel pretty uncertain about the idea of having extremely brilliant and dedicated men and women, who also happen to be extremely frustrated and desperate. (And let us admit it, the fact that they have been training for years on the handling of sharp tools does not help)

So what do we do? Assuming we won't be needing neurosurgeons, and stopping to train new ones, is a risky bet. If there are no robots to replace them in 15 years that will create a real problem. But if we assume that the robots won't exist, we may end up training people for a job that won't be available when they finish their training. And of course this problem is not limited to neurosurgeon. It applies to any precision work that requires a huge amount of knowledge and technicality. Two domains in which automatons excel. From piloting a plane to drilling holes into high end motor parts, from mixing unstable chemicals to growing crops, from holding accounting books for a big company to managing the delivery of goods all over the world.

The easy solution would be for lawmaker to ban the replacement of humans by automatons. India's transport minister has already vowed to ban self-driving cars to protect jobs . This approach may work for a short while, but it will turn into a poison for countries that choose this path. First of all it will have a huge negative impact on the economy and the prosperity of said countries. Automatons will provide a massive boost to health, productivity and stability.

Self driving cars will take jobs from humans, but they will provide faster and safer travel, optimized energy management, improved delivery of goods etc. Moreover for the "ban automation" approach to have any noticeable impact every form of automation will have to be banned. Saving jobs just in the transport domain won't help – automatons will have to be forbidden in the medical domain, in factories, in public services, in education... In every aspect of the inner workings of a country, the automaton ban will translate into a handicap.

On the contrary, embracing automation can be a formidable boost for growth. First of all the cost of labor disappear. That means occidental countries can create industries, even heavy ones, and still be competitive. The demand for robots and robot parts is going to grow. The need for better gears, better lubricants, more efficient motors and batteries will rise. Substantial financial gains awaits the companies that will lead the market in these domains. Algorithms and optimization may not be a direct source of revenue on their own. (In most countries they cannot be patented unless they are tied to an industrial process.) But teams of scientists with extensive knowledge on these technologies will be a formidable asset. Both for states which train them and firms which hire them.

Growing up with robots.

Unemployment will rise. Not everyone can be an automation scientist. This last industrial revolution will not just free the work force from a type of unqualified work. It will make all technical aspect of work inaccessible to humans. In previous industrial revolutions, innovations generated a shift in available jobs. New domains appeared and new tasks were created. This time we can assume that these new tasks will be accomplished by automatons. Actually they will be thought with automation in mind. Automatons will fill current positions, allowing for new jobs to be created, jobs that automatons will also fill.

Given all this, there is a real (and quite urgent) need to rethink job market, or to be more precise unemployment and education. Education is on the verge of becoming a lottery. In a world where entire domains can disappear due to automation, the safe bet is to have a broad and generic training. And of course the worst option would be specialization. Specialists are limited to one field of expertise, investing years of training and practice into their skills. If said field becomes compatible with automation, their expertise is useless.

But as a society we need specialists to be an elite. Someone who had to become a surgeon because he was last in medical school is a scary thought. Likely we would not want the people who operates nuclear power plants to be average students.

We need to find a way to motivate the best elements to train for the most complex position. We also need to account for rising unemployment. Preferably we should do all this without creating ghettos or segregation.

P.S : After writting this article, I found the following video :

Made by CGP Grey

Both subject and conclusion are the same, yet the analysis is different. And there are some points of disagreement - for example, I do think that creators (artists, problem solvers, author etc.) will keep their jobs for a while. The first wave, in my opinion will mainly impact technical jobs. Still a very good video to watch.

Illustration Picture : An artist's concept portrays a NASA Mars Exploration Rover on the surface of Mars - Credits : NASA/JPL/Cornell University - URL