David Brooks: Volition, Not Intelligence, Wins the AI Age

David Brooks used his latest Atlantic essay to make a claim that cuts against the standard 'AI will make smart people smarter' storyline. His argument is simpler and a little more uncomfortable: what will separate people in this era is not how smart they are but their relationship to mental effort.

The line worth writing down is his guiding principle: 'When intelligence is plentiful, volition is valuable.' If a model can answer any question, the differentiator moves upstream, to who bothers to press it, argue with it, or use it to do harder work rather than less work. Brooks sketches two dispositions on either side of that fork. The people who make a difference, he writes, are not the ones who 'seek relaxation and passively use AI to work less' but the ones who 'seek improvement and actively wrestle with AI to develop their own mental capabilities and accomplish more.' On the other end are what psychologists call 'cognitive misers,' people who find hard thinking unpleasant and take any opportunity to avoid it. In between sit most of us.

The essay is a values argument, not a data one, and it reads accordingly. Brooks leans on a psychological trait ('need for cognition') that has decades of research behind it, and stretches it into a claim about who will thrive in an economy suddenly full of cheap cognition. Two tracked AI experts in our directory shared the piece, which is a modest indicator that operators are treating it as a lens to argue with rather than a data point to cite.

Brooks does not attach the argument to any particular company, product, or workforce dataset, so treat it as a frame rather than evidence. He also does not say what schools, managers, or AI product teams should actually do about the split he describes, only that it is coming.

Still, the operational read is real. If he is right, 'does this person lean into difficulty' becomes a management signal that matters more than credentials, and AI products that quietly encourage friction, that push back, ask better questions, and make the user do some of the work, start to look more valuable than the ones that race to hand over an answer. The people who treat the model as a tool to think harder will pull away from the people who treat it as a way to think less.