The writer is the global co-head of investment banking at Goldman Sachs.

Data is the foundation of the artificial intelligence revolution, but AI is also revolutionising the market for data. Developers are racing to invest billions of dollars to build the infrastructure to power vast AI systems. That rapid expansion has led to a surge in demand for data, creating the potential for companies to generate significant economic value.

AI systems are typically described as having three main components — power, compute and data. These refer to the electricity required to power data centres, the chips needed to conduct computations at mind-boggling speeds, and the data necessary to train AI models. Of these critical components, it is data that is least discussed, perhaps because data centres and semiconductors are physical things you can see and touch. (It’s admittedly difficult to hold up a data packet during an onstage keynote.)