It’s common for uncertainty to be perceived as a problem to be solved, a gap to be closed. How many times have you apologised for not knowing the neat, exact answer to something? How often have you been praised, or given praise, for certainty (even if you weren’t, actually, completely certain)? Reaching an answer, finding a solution, has come to mean relief from the discomfort of not-yet-knowing. The destination we have finally arrived at, that we can sink into like a warm, deep-cushioned sofa after a winter walk. The end.

I want to make the case that we've made, and continue to make, a massive mistake by focusing on the answer rather than the question. Not being certain should never be the state we rush through, reaching for the threads of something – anything - concrete and final. Instead, let’s redefine it as a state of thinking. Of contemplating, where ideas can meld and spin and thread themselves together into new and surprising combinations. What if, far from being a muddled middle we leave behind as fast as possible, uncertainty is not just desirable but actually the whole point of our complicated, messy human brains?

Our neurochemical reward systems are not built for certainty. Dopamine has been described for decades as the chemistry of pleasure, the payout you feel when you finally land on the prize, solution or answer. But that isn't what the evidence shows. Scientists at the University of Michigan spent years studying the difference between wanting and liking, comparing the pull towards something with the pleasure of having it, and found that dopamine drives the wanting, while the liking runs on smaller, more fragile brain systems. Dopamine is the provisions we give our brains for the journey, not the reward for reaching the destination.

Moreover, when researchers looked specifically at curiosity, the reward circuitry engaged while people tackled a question they couldn't yet answer, and that curious state improved their memory for the answer when it eventually came. In your brain, a question you care about and haven't answered isn't a discomfort, it’s a state of reward. Why, then, the itch to solve? If our brains are happy with answers being open. Why aren’t we?

I believe the reason is cultural, not biological. The result of a system that reaches for ever more efficiency, ever faster resolutions, sometimes – perhaps often – at the expense of depth and value.  And now, with LLMS, we've built machines whose entire nature is to abolish uncertainty for good. AI models can collapse anything open into resolution in seconds. It answers questions you hadn’t even arrived at yet, makes predictions about the next solution, and the next, and the next. AI models have no state labelled ‘not-knowing’; only the most probable completion. So, uncertainty is the one thing it structurally cannot do. But in not allowing uncertainty, it also does not allow that rewarding state of our brains. By the way, I can’t find evidence for this (yet) but I believe this is why the infinite scroll is so addictive, trapping us inside the ‘discovery’ state like a kind of modern day island of Ogygia, never delivering the pause that will allow us to escape the reward.

Keats gave a name to the ability to remain in uncertainties, mysteries and doubts, without any irritable reaching after fact and reason: negative capability. Not landing on an answer because you've been able to reach the best one, but because you can't bear the uncertainty any longer. That irritable reaching is exactly what the LLMs frictionless answer engines formalise and industrialise, making the finding out effortless. So we never struggle, and never find out what might have happened if we sat with the negative capability a little longer. As a poet, Keats would have understood that deeper levels of meaning don’t come easy. There is no ‘hack’ for understanding, even if we’ve landed on – or been given - answers.

I believe the things most worth thinking about are uncertain by nature. To risk labouring the point; that's what makes them worth thinking about! The questions that matter are tricky, messy, unresolvable. The real work – the human work – is to keep them open long enough for our brains to feel the reward of wrangling. Certainty feels like an elaborately wrapped gift, especially in today’s stressful, time-poor world. But all too often the answer is an empty box. What we need, it turns out, is something like a knot in the gift bow that takes effort to work out how to unravel. That’s the real gift; the uncertainty before we know what’s inside.