Planes Aren’t Birds but Both Can Fly. Computers Aren’t Brains but Both Are Intelligent.

A long time ago —“Oh, look at that bird. Why can’t we fly like that?” Today —we are flying like birds, using planes, helicopters, rockets, wingsuits and many more of our inventions. We fly higher, faster, and for longer. And it all started with kites. That’s when we started, unknowingly at first, to harness new sets of rules. The rules of flight.

The explanation of why we can build anything at all, derives from one of our best theories of how the universe works. It goes as follows. At the scale of human beings, the universe is made of systems that are made of parts, that when arranged and related to each other in specific ways, lead to specific behaviours. These parts and relationships can be rearranged to display different behaviours. Also, the same behaviours can be displayed by two systems made of different parts and different relationships. Whether a behaviour can be displayed by a given system depends on the rules it is following, which at the most fundamental level are the rules defined by the laws of physics in the language of mathematics.

We know that this theory works because we have built all kinds of things in the past. We tend to become curious about behaviours that we can observe and start building things to imitate them. For example: we could observe what it is to fly but couldn’t explain why it worked and how it worked. We figured it out by building flying objects —kites at first, followed by sky lanterns, primitive rockets, man lifting kites, hot air balloons, planes, more rockets, etc. Over time we figured out many more of the rules that needed to be followed to fly faster, higher, and for longer.

We are currently doing the same for intelligence. We know what an intelligent person is like, because we can observe them, but we can’t explain why intelligence works and how it works.

Intelligence is a skill which allows us to figure out how to achieve a particular outcome. An intelligent person can achieve particular outcomes in highly complex scenarios, in which a less intelligent person wouldn’t be able to do so. In other words, more intelligence can solve more complex problems.

We started creating intelligent humans through teaching, using movement, spoken and written language, and numbers. More recently we started making intelligent tools —computers— which we use to replace and improve some of our own intelligence. We program intelligence into these tools but are now at a point where we can train them and make them evolve into intelligent tools. The intelligent behaviour displayed by these tools is the same as the intelligence displayed by a human brain.

Let’s consider flight again. What would a person living in the days before human enabled flight have imagined about the use of flying? They probably would have imagined doing things that they couldn’t have done without flying.

We are doing the same with intelligence. We already do and will continue to use new and more intelligent tools to do things that we couldn’t have done before, for example, solving problems which are considered unsolvable today. We will increasingly intervene, to change and to manipulate the world. More doing will lead to more change.

If you like change, the future promises to be very exciting. Nevertheless, there is one reason that more change wouldn’t happen, namely, if we decide not to build intelligent tools. At the same time, it is guaranteed that there will be intelligent tools if we build them.

How intelligent will these tools that we build be? We figured out how to make things fly and how they could outperform birds, for that we had to figure out the underlying rules of flight. The same might happen for intelligence. Computers are already outperforming humans in some tasks that require intelligence, but as long as we don’t know the exact rules underlying it, we can’t predict how intelligent our tools will become and what intelligent behaviours they will display.