What is intelligence?

There is often confusion and conflation around the term intelligence. Many of us fall into narrow definitions. Broken definitions, even. By broken I mean a definition that gets mired in confusion; one that isn’t good enough to make sense of a modern world where many kinds of agents display many different kinds of intelligence. Such broken definitions are often too specific or too arbitrary. Some are rooted in false dichotomies. Saying they “miss the point” would be an understatement – the problem with such broken definitions is they aren’t disembodied things. They roost in people’s minds. If someone can’t conceive of intelligence as being independent of human cognition, we have a problem.

Many of our current language patterns are liabilities. These emerge sometimes in organizational cultures where hazy definitions slide around and people are wary to admit their uncertainty about what others mean, much less their own ignorance. Sometimes it feels like a big charade where no one wants to hurt anyone’s feelings nor appear uninformed. And so it goes, some kind of elaborate mystical ritual where the confused participants lead each other further into madness.

With this in mind, I find great value in Stuart Russell’s definition of intelligence: the ability of an agent to solve some task. An agent is anything that makes a decision: a human, an animal, a system of any kind. This definition intentionally leaves out any notion of (a) humans; (b) consciousness; (c) some arbitrary quality line.

This usage bypasses a lot of traps that mire down other conceptualizations of intelligence. I recommend finding a way to shift conversations towards it wherever possible.

One might ask if Russell’s definition merely “kicks the can down the road”. I don’t think so. This is just a word game. It reduces the scope and relaxes the pressure to generalize. This reframing helps by encouraging people to define their metrics for solving a particular task well. It is one step in the right direction that offers two big benefits. First, it helps us to stop pretending like we all know what each other means. Second, it moves us towards answering tractable questions. Incidentally, I think the definition is compatible with how a lot of organizations solve problems. Leadership can set the overall direction and can delegate accordingly.

Ok, but what about “general” intelligence you say? Good question, but let’s take this one step at a time. Wait until a group of people have put in the work and demonstrated some ability to reach some consensus. Defining intelligence even for particular tasks may not be easy, depending on your stakeholders. Defining general intelligence is harder and can be contentious. It often becomes a lightning rod for all number of other disagreements.