What can children teach us about artificial intelligence, and what can AI teach us about human development? When people talk about AI, it’s easy to slip into the idea that “intelligence” is one thing you either have a lot of or a little of. Alison Gopnik (UC Berkeley) argues that there isn’t one single, all-purpose intelligence. Instead, there are different kinds of thinking that involve real trade-offs. One way to see those trade-offs clearly, she suggests, is to look at how humans think across a lifetime.

Gopnik uses a classic idea from computer science, the explore–exploit trade-off, to explain what changes as we grow. Exploitation is the kind of intelligence that helps you set a goal, make a plan, and take action to achieve it. Exploration is different: it’s less about reaching a goal and more about learning how the world works by trying things, gathering information, and staying curious. In Gopnik’s view, childhood leans heavily toward exploration, and that’s a feature, not a bug.

She also discusses later life, when people often play larger roles in care and in passing knowledge across generations. Here, Gopnik introduces a third idea from AI research called empowerment. She describes empowerment as strengthening the link between what you do and what happens next, while also keeping a wide range of actions available. Put simply, it’s about finding parts of the world you can influence, where your choices can genuinely change outcomes. Across childhood, adulthood, and elderhood, Gopnik argues that these different strengths—exploring, exploiting, and empowering—work together, and that combination helps explain human success.

Learn more about the different types of intelligence across a lifetime. Watch Three Ages and Three Intelligences: Exploit, Explore, Empower with Alison Gopnik.