That’s Pablo Picasso, being interviewed about computers in mid-1960s Paris. The machines he was talking about bear almost no resemblance to what sits on your desk today - but his observation is more accurate now than it was then. The most powerful AI systems ever built are extraordinary answer machines. They are not, and for the foreseeable future will not be, question machines.

That distinction matters more than almost anything else I say on stage.

Because the skills that will define human value in an AI world are not the ones we’ve spent the last century building education systems around. They are not the ability to recall information, process data, or execute repeatable tasks. Machines do all of those things better than us now, and will do them better still tomorrow. The skills that matter are the ones that sit in the space machines cannot reach.

I think there are three. I’ve thought about this for a long time, and I want to explain not just what they are but why - because the reasoning matters as much as the conclusion.

Creativity

We are already facing problems we have never encountered before. Climate. Health. The redesign of economies and institutions for a world that technology is reshaping faster than our organisations can adapt. The defining characteristic of these problems is that past experience is an unreliable guide to solving them. They require us to think in ways we haven’t thought before, to make connections that haven’t been made, to imagine possibilities that don’t yet exist.

Generative AI cannot do this. The entire premise of the technology is that it reflects what we’ve already said and done - it rhymes with reality rather than creating it. It is a mirror, not a window. The creativity we need for the hardest problems ahead is the kind that looks through the window, not into the mirror. That is irreducibly human, and we are not developing it nearly fast enough.

Empathy

In a world of dark, cold logic - of algorithms and optimisation and statistical likelihood - the ability to understand and share the feelings of another person becomes not a soft skill but a strategic one.

Machines are busy crunching numbers. They will get better at predicting what we feel. They will not get better at actually feeling it alongside us, or knowing what to do with that in the complexity of a real human relationship. The border between human and machine runs, in large part, through empathy. It is the quality most at risk of atrophy in a world where screens mediate more and more of our relationships - and the one most worth actively developing.

Accountability

This is the one people underestimate, so let me make the case plainly.

Just because the algorithm gives you an answer doesn’t make it right.

Consider two datasets: the change in average global surface temperature over time, and the number of pirates operating in the high seas. Plot them together and you get a near-perfect inverse relationship - as pirates decline, temperatures rise. The data, faithfully processed, suggests that if we want to address climate change, we should perhaps consider more pirates.

This is funny until you realise that the same basic error - confusing correlation with causation, trusting the output without interrogating the logic - is happening in boardrooms, clinical settings, and policy decisions every day. AI systems produce confident, fluent, plausible-sounding answers. The human skill we need is the will and the judgment to ask whether those answers are actually right, and to retain ownership of the outcome regardless.

These three were not chosen because they sound reassuring or because they give us hope for a more human future irrespective of what technology does. They were chosen because they are the precise complements to what machines do well - and because, despite what Hollywood tells us, they will remain fundamentally human for decades to come.

The question isn’t whether you have them. The question is whether you’re developing them - and whether the organisations and education systems we’ve built are helping or hindering that development.

Most aren’t. Not because people don’t care, but because we’ve spent so long rewarding the skills that machines now do better than us. We built the wrong curriculum, and we’re still teaching from it.

The rise of the humans starts here.