The science behind tomorrow’s economy

Science matters. The Industrial Revolution marked the start of the most dramatic acceleration in economic growth in human history, but it was only possible because of the Scientific Revolution that preceded it. Systematic observation, experimentation and the sharing of knowledge displaced superstition, word of mouth and secrecy as ways of understanding the world.

From the steam engine to AI, every innovation required brave souls who were prepared to invest their capital, and even more importantly, their careers in fields whose potential was far from certain. Hindsight celebrates them as pioneers, but forgets their peers who made the wrong bets. Finding the next great opportunity is just as difficult today as it was then. There is an astonishing breadth of scientific research and early-stage experimentation underway around the world and somewhere amongst this quarry of ideas lies the foundation stone for the next great engine of human progress.

Given the world’s focus on AI, it is easy to overlook the breadth of extraordinary scientific research and discovery happening at the same time. Scientists are learning to engineer living organisms, manipulate matter at a molecular level and explore nuclear fusion energy for the future. Engineers are working with scientists to build robots capable of operating inside the human body while others are looking to manufacture everything from new medicines to high performance fibre optics in space. Any one of these and more could trigger the next revolution if there are pioneers who are brave enough to invest either their money or time today.

These are the so-called deep tech opportunities, as revolutionary to us today as steam locomotives, electricity, telegraphs and the internal combustion engine were to pioneers of their time. But, it is hard to pick both the science that shows the most promise and the timing when it can be realised. The deep tech landscape is littered with opportunities that have been tantalisingly within five years for decades and may never come to fruition. I’ve written about this topic before, but the amount of money going into AI today, paired with major opportunities and challenges in other fields of deep tech, makes it worth taking a look again.

We should perhaps start by staying close to the digital and AI ecosystem and look at what might be the next revolution, quantum computing. I’ve been writing about quantum advances since 2014, and has been the subject of both hype and scepticism dating back to the 1990s. If quantum computing fulfils its promise, it could overcome some of the constraints of digital computing, potentially including those limiting AI, both in the problems it solves and possibly even the energy it needs. However, the breadth of fundamental physics research being funded under the banner of quantum computing makes me wonder whether some physicists are being tempted to brand their projects as computing investments to attract investors, instead of presenting them as pure research that depends on university and other science grants.

Moving only slightly afield, mathematics is one of the fields most disrupted by AI. Often science and maths play tag with engineering, as happened with the steam engine, which was developed by engineers, improved through scientific experimentation and subsequently better understood and refined through the maths of thermodynamics. Similarly, Faraday connected magnetism with electric currents, Maxwell developed the equations that governed them, Hertz demonstrated the electromagnetic waves they predicted, and Marconi turned those discoveries into radio communications. We can expect that some of the most far-reaching deep tech opportunities are hiding in the reams of draft papers being generated by frontier AI companies and awaiting verification by mathematicians.

Advances in AI and maths are already being combined to improve protein structure prediction and its many applications from pharmaceuticals through to bioengineering. With mass extinctions threatening potential drug discoveries in nature, synthetic approaches to developing new drug candidates are increasingly complementing fieldwork. Similarly, bioengineering may be critical to the recovery of our planet with engineered organisms being developed to tackle everything from breaking down plastic pollution to reducing our dependence on expensive and polluting fertilisers.

Materials science is similarly being turned on its head. Instead of studying existing materials to find new applications, scientists are using a combination of AI and maths to design materials at a molecular or structural level and predict their properties for useful applications before the chemists even step into the lab.

Major advances in technology are usually the result of the convergence of progress in different fields rather than a lone genius coming up with an idea. Arguably the GPU-fuelled AI boom was accelerated by Mark Zuckerberg’s push for compute to fuel his metaverse vision, recommendation engines and AI ambitions, demonstrating that companies can commit billions to technologies whose application turns out to be quite different to what they originally envisioned. Similarly, autonomous cars have leapt forward in capability through the convergence of navigation tools with new sensor technology and AI-driven control systems, which have contributed to physical AI and a new wave of drones, robots and autonomous vehicles.

One of the consequences of the boom in AI is an enormous demand for power to fuel the data centres, far beyond what energy companies anticipated. As a result, there is an innovation race between those who seek to improve renewable energy such as solar and wind, and those looking to reinvent nuclear fission or unlock new sources of geothermal and even fusion energy. Equally important are the advances needed, and increasingly realised, in energy storage to support the stability of a growing grid and the mobility of increasingly autonomous machines. With large-scale demand comes capital to invest and previously insurmountable engineering challenges become commercially worthwhile to tackle. What environmentalists have struggled to achieve through advocacy, perhaps science can realise through market imperatives.

There is no magic formula to predict with certainty where advances in science will intersect with the demand for new technology. Even Einstein could hardly have anticipated general relativity would one day enable satellite navigation. But understanding the trends and where investment is likely to flow helps to put current research priorities into perspective and encourage collaboration between those working on different parts of the puzzle. Governments, universities, researchers, investors and businesses all need to find ways to plan and work together if we are to realise some of the most exciting opportunities in science. While it is unreasonable to expect certainty today, hindsight will make the next decade a fascinating period of history.