As a simple example, computers can now fully assist you to do your taxes, sometimes better than a human could, but the accountants are still around, working roughly 40h a week and they are in high demand. Computer programs can improve their job but can’t replace the human approach, adaption to specific personal situations, and they can’t innovate beyond what was initially hard-coded.

Yes, AI can, or at least will soon be able to drive a car autonomously, but we will still need truckers. Why? Think about it, trains or subways are on tracks, it would have been pretty easy to hard-code their shift, but we still have train and subway operators. Even pilots can fly airplanes today without any manual intervention, and this happens even without AI technologies. The role of pilots has been completely transformed, but pilots are still needed. We are still a long way from trusting these lines of code enough to fly an unmanned aircraft.

AI technologies allow computers to perform new tasks that hard-coded computer programs couldn’t perform before. With Computer Vision, Natural Language Processing, Reinforcement Learning, etc., tons of new applications can now make many of our tasks easier. These technologies are truly impressive, but to date, their applications are still very narrowed. Although every new publication tends to say that we are getting a step closer to a General Artificial Intelligence, to drive hype and increase click-through rates, we are still far from a real artificially intelligent being, and even once that happens, we will still need people to make the right decisions, to maintain and manage their software, to research and improve them and to have this highly needed human approach.

I think almost every industry will be impacted by AI in the near future. These technologies help us build new tools that we can use to free up people from repetitive tasks, optimize processes, facilitate monitoring, support creative and innovative work, improve efficiency, and much more. Every area will benefit from those new technologies in one way or another, but some will certainly take more time to be affected than others.

The sectors affected today are certainly those where data to train these algorithms is readily available. That’s why Alphabet (Google), Meta (Facebook), and Amazon are behind some of the most transformational applications. Then come those where data can be easily collected and the expected behaviors are more or less circumscribed (ex. Obey the rules of the road, don’t crash, get me from point A to point B for self-driving cars). Later, these technologies will aid decision-making in medical roles, but they won’t replace the need for the physician’s or psychologist's human approach. Surprisingly, some of the toughest jobs for AI to improve might be some of the manual labor involved in construction - plumbing, for example. Some applications can help improve their work, but even with the advance of AI in terms of software, we still have a long way to go to develop mechanics that would allow a robot to go to the right place, squeeze into tight areas, find the right tools and show enough creativity to adapt to a specific problem.

In a nutshell, I don’t think we should fear the impact of AI on any of our roles, but we should all educate ourselves so that we can benefit from these technologies to improve and facilitate our work.