Superhuman Problems We Can Finally Solve

The conversation about AI is stuck. Every week brings another article about chatbots replacing copywriters, another survey about job anxiety, another breathless take on productivity gains. And look, those things matter. But they are a sideshow. The real story is bigger, harder, and far more exciting than anything happening in your inbox.

The wrong conversation

Here is what frustrates me about the current AI discourse. We are having a debate about whether AI will write our emails faster, while AI is quietly solving problems that the whole of human science could not crack. The framing is wrong. Not because productivity does not matter, but because focusing on it is like watching the Wright brothers take off and asking whether it will speed up the postal service. Technically true. Spectacularly missing the point.

The problems that define our century are not small. Climate change, drug-resistant disease, the pace of medical diagnosis. These are superhuman problems. Not superhuman because they require genius, but superhuman because they require processing complexity that no number of brilliant researchers could ever handle alone. They are too expensive, too slow, and involve too many variables for conventional approaches.

What is actually happening

Consider drug discovery. Getting a single new drug to market used to take over a decade and cost upwards of a billion pounds. DeepMind's AlphaFold predicted more than 200 million protein structures, work that is now used by over 3 million researchers across 190 countries. It won the Nobel Prize in Chemistry. That is not a productivity hack. That is a step change in what humanity can attempt.

Or take climate. Google DeepMind's GNoME tool identified more than 2 million new crystal structures with potential applications in renewable energy storage. That is 45 times the number science had previously catalogued. AI weather models now outperform the best physics-based forecasting systems, and they do it with a fraction of the computational power.

In healthcare, European researchers have built AI models that can predict over 1,000 medical conditions, including cancers and heart attacks, more than a decade before a formal diagnosis. In the UK, AI-powered tools are already reducing prostate cancer diagnosis wait times by analysing imaging scans that would take human specialists far longer to process.

Can AI help solve climate change?

Yes, and it already is. AI is optimising renewable energy grids in real time, predicting supply and demand fluctuations that humans cannot model fast enough. Google Maps' eco-routing feature alone has prevented over 1 million tonnes of CO2 emissions annually by suggesting fuel-efficient routes. AI-enhanced climate models are cutting the cost of running ensemble forecasts without sacrificing accuracy. The question is no longer whether AI can help. It is how quickly we can point it at the problems that matter most.

What this means for you

If you lead an organisation, the strategic question is not just how AI improves your operations. It is whether you are paying attention to the industries being reshaped by AI solving problems that were previously unsolvable. Drug discovery timelines are shrinking by up to 70%. New materials are being found at 45 times the historic rate. These shifts will reshape supply chains, regulatory landscapes, and competitive dynamics within the decade. The AI Leaders Fellowship exists precisely to help senior leaders think at this level.

The bigger picture

I spend a lot of time talking to leaders who are worried about AI. And worry is reasonable. But I want more of those conversations to start with this: we are living through a moment where problems our grandparents assumed were permanent are becoming solvable. That is worth getting excited about. If you want to go deeper on why AI is not the threat the headlines suggest, this talk is a good place to start. And if you want to understand what this means for your own leadership strategy, that is what we are here for.