Why Wondering is Important in the Age of AI

Curiosity is not just a feeling. It is a cognitive state that changes how the brain processes and retains information. Research in learning science has shown consistently that when people are in a state of genuine curiosity, when they are actively wondering rather than passively receiving, they form stronger memories, make more creative connections, and are more likely to transfer what they learn to new contexts. There is something almost physical about the experience of genuinely not knowing something and wanting to. The mind leans in. It becomes porous in a way that it is not when it is simply confirming what it already suspects.

The problem with instant answers is not the answers themselves. It is that they can short-circuit the lean. When I type a question into an AI and get a fluent, confident, well-structured response in three seconds, my brain registers the question as closed. Filed. Done. I move on. But if the same question had sat with me for a day, had bothered me over lunch and come back to me in the shower, I might have arrived at something different. Not necessarily the correct answer, but a richer set of associations, a more personal relationship with the idea.

This is not a romantic argument against technology. I use AI constantly, and I find it genuinely useful. But I have started paying attention to the difference between questions I should ask a machine and questions I should sit with myself first. The former tends to be informational. The latter tends to be generative: what do I actually think about this, what is the most interesting angle nobody is talking about, what does this remind me of that might not be obvious. The generative questions need the friction. Feed them to a machine too quickly and you get back something competent and slightly hollow.

Here is a tension I find genuinely interesting. We are living through a moment where the quality of your questions matters more than it ever has before, precisely because AI has made answers so cheap. If everyone has access to the same information at the same speed, the differentiator is no longer who knows more. It is who is asking better questions, who is framing problems in more interesting ways, who is noticing the thing that has not been asked yet. That is a premium on wondering. And yet the very tool that creates that premium also, if we are not thoughtful about how we use it, gradually atrophies our ability to wonder well.

Good questions do not arrive fully formed. They develop through a process that requires spending time in a problem space without immediately resolving it. You read around an area. You notice what does not quite fit. You feel the slight irritation of an explanation that almost works but leaves something out. You follow a tangent that turns out to be a dead end, and on the way back you notice something you missed before. That messy, non-linear, inefficient process is how genuinely interesting questions get formed.

I have been thinking about this in the context of how I approach my own work. When I am deep in a problem I care about, I have started deliberately delaying the moment when I ask AI for help. Not because the AI answer will be wrong, often it will be quite good, but because I want to have spent enough time wondering that I know what I am actually asking. The AI becomes more useful, not less, when I bring more of myself to the conversation. Wondering, it turns out, is not just a pleasant pastime. It is preparation.

There is a version of the future I find hopeful and a version I find troubling, and the difference between them comes down almost entirely to whether we remain serious about cultivating wonder. The hopeful version is one where AI handles the informational work so well that human minds are genuinely freed up to do the deeper, stranger, more generative thinking that machines do not do. Where having instant access to any fact means that facts are no longer the bottleneck, and what matters instead is the quality of the framework you build around them. That version is available to us. It requires intention.

The troubling version is the one where convenience quietly colonizes the entire thinking process. Where every moment of not knowing is immediately resolved, where wondering is treated as an inefficiency to be eliminated, where the muscle of sitting with uncertainty gradually weakens from disuse. Where we produce and consume enormous quantities of AI-assisted thinking that is technically correct and thoroughly derivative because nobody involved spent enough time in genuine uncertainty to generate anything that surprised them. That version is also available to us. It is, honestly, the path of least resistance.

So here is what I keep coming back to, lying awake and wondering about things the way I did as a kid, only now with the faint glow of a phone on the bedside table that could answer any of it in seconds. The question is not whether AI will keep getting better at answering things. It will. The question is whether we will stay curious enough, disciplined enough, and a little stubborn enough to keep wondering before we ask. To let the question breathe. To notice what we think before we find out what the machine thinks. Because wondering is not just what humans do instead of knowing. It is, very often, how the best knowing eventually happens.