An algorithm is a process or set of rules to be followed in calculations or other problem-solving operations. Basically, an algorithm is a set of step-by-step instructions for solving a problem. There are simple algorithms, like how to tie a shoelace, to very complex algorithms, like those that are used to manage stocks for investors.
The algorithms driving self-driving cars are incredibly complex. They take inputs from a variety of sensors and recognize the objects, such as cars, pedestrians, and road signs. The objects are then compared to the a library of sorts that the computer has learned from Machine Learning and its actions are predicted. The computer then takes all this information and decides the best course of action for the car. This is done many times every second to ensure total control of the situation.
These algorithms are incredibly powerful and are constantly being researched and updated, but there is always the possibility of a situation arising that the car is unprepared for. So what should be done when an algorithm causes an accident or fails? Is the car manufacturer at fault? Or the coder? Or do we just add this edge case to the system and move on?
Another potential problem with these algorithms is the issue of power and control. By putting our trust in these algorithms, our society is putting our trust into a single entity, as opposed into the million collective minds of our own. This collection of power can have social consequences, such as when Facebook’s censorship algorithms curated the news.
These algorithms are incredibly complex and well executed. But we are putting a lot of trust into something created by our flawed selves, so we should make sure that we are creating something better than ourselves.