THE TRAIT THAT PREDICTS AI SUCCESS

Some people meet a hard problem after lunch and happily lose the afternoon to it.

Others – brilliant people among them – will avoid the same question for a week.

You work with both.

Psychologists have been measuring the difference since 1982. The trait is called ‘need for cognition’. It measures your appetite for effortful thought.

And AI has changed what that appetite is worth. It might, in fact, be the most important factor in successful AI use.

In a recent essay for The Atlantic, David Brooks sketched three ways it plays out.

Productive passengers find hard thinking unpleasant and take any chance to avoid it. They use AI to think less and produce more. Often that results in AI slop.

Reluctant optimisers will put in effort when they really care, but don’t intrinsically enjoy it. They mean to keep thinking, but deadlines and fatigue often win.

Mental marathoners enjoy thinking hard and use AI to think harder – bigger questions, stranger territory, work they'd never have attempted alone.

These hungry ones make the machine argue back: attack the plan, find the weakness, spoil the neat conclusion. They give it the donkey work and keep building their judgement. And when it saves them an hour, they point it at something harder.

It turns out how much you gain from AI may depend not so much on how smart you are - but on how hungry you are to think.

So in-the-end-at-the-end?

Angela Duckworth famously argued that effort counts twice: it turns talent into skill, then skill into achievement. And AI seems to make it count yet again - by multiplying the effort you bring to it. Because appetite is something digital intelligence can't supply.

The colleague producing AI slop probably isn’t a hungry thinker.

So the question becomes: how do you spark then feed that appetite -  to build an organisation of cognitive marathon runners?