On Intelligence

If we think seriously about intelligence for a little while, we recognize that it isn’t just a random thing in the world, a trait like other traits, but rather a genre of order in the universe. At its most immediate, intelligence is supposedly the way we human beings have evolved to survive in the world: as we like to think of ourselves, it is our intelligence, as opposed to other traits like, say, raw speed, viciousness, stealth, sheer profligacy, or the ability to photosynthesize, that other organisms may have employed. To frame intelligence thus frames it helpfully in two dimension: it is relational, and it has evolved in an environment of relations for a very long time.

Adaptability isn’t the same thing as intelligence, but the ideas are clearly related. The obvious benefit of our human intelligence, we like to think, is our ability to think strategically and creatively in the face of various challenges, to evolve solutions to dynamic problems over time. But other animals we think of as ‘less intelligent’ – dogs, cows, fish, birds – are also quite capable of managing their adaptation to dynamic challenges and opportunities. So intelligence, if we can locate it in a kind of rapid-response strategic behaviour, isn’t something that is unique to us humans: rather it is pretty apparent throughout the animal kingdom, and not unimaginable even in plant life. And if that’s true, then to look to specifically human facets of intelligence – like vocabulary and grammar – is missing quite a lot.

Relationality is foundational here. Intelligence evolved, presumably, as a strategy for an individual to manage its relationship with its environment: things like food sources, security threats, allies, mates, friends, hygiene, fun, snowstorms, and sunsets. I remember learning that the tiny amoeba swims toward areas of safe pH levels and swims away from unsafe ones; such behaviour might be somehow ‘hard-coded’ in the protein structures of its tiny body, but surely such relational behaviour is a kind of prototype for the whole thing we do with our brains.

If intelligence is an evolved strategy for interacting with one’s environment, then to look at it as merely a function, as we do with LLMs, is to really miss the point. The thing this misses is often gestured to in the current discourse as “intentionality,” the full meaning of which is probably lost on most of us without formal philosophical training – what I am getting at here is the “aboutness” of thoughts, the “holding” of beliefs, the trustworthiness of facts. These are all relational concepts. Representing knowledge in a form outside a person’s head (one of our great human tricks, whether it be cave paintings, linguistic utterances, or computer programs) is not the same thing as intelligence. To build on that, “generating” written descriptions or images – representations of knowledge – does not amount to intelligence either. It’s a clever and probably useful trick, and arguably an important levelling-up on static representational forms. But it falls so incredibly short of “intelligence” that… really people, we have to stop using that word.

All of our clever representations of knowledge, ideas, facts and models about the world, can only have value if they are for someone. Without that someone, they are merely inscriptions: they require a reader, an interpreter, someone with an intentional relation and a dynamic trajectory in the world for them to have value.