WHO CREATES DEMAND FOR AI?

Digital intelligence is slowly, surely becoming part of the operating system of the economy.

And Big Tech's projected $7.6tn bet between now and 2031 is one heavily capitalised side of the equation. The supply side, to be precise.

On the other side sits diffusing the technology into the fabric of organisations. What Dario Amodei calls the 'other exponential'. And what we might call: The demand side.

It’s economics 101 to know that demand has to meet supply. And usually, we think of the demand arising first - creating the conditions in which supply can expand.

But so confident are the frontier labs - and hyperscale data centres on which they rely - that demand will arrive, supply is being planned ahead of demand. A lot of it.

So what? Well, as one AI commentator put it: only AI training can save the economy. Humans are the ones generating the demand. And without it, spending trillions in capital is going to look like quite the misallocation. It’d be like hosting a very big party. But forgetting to invite the guests.

So how might we nudge this along?

We’ve found that being joyously specific about how teams can integrate AI into their work - through role-specific coaching and practice - is very effective. Spend time with them. Help them make some magic happen. You’ll find that self-fulfilling curiosity and enthusiasm quickly take hold.

Make this visible across the organisation and AI becomes a focal point. Which means that suddenly, everyone wants in.

There’s not a lot of evidence that the frontier labs and hyperscalers understand the success of their mission is dependent on finding effective training methods. But from where we’re standing, it could be a determining factor.

Will those trillions be mispent? Or will there be a rip-roaring party we can all enjoy?

The quality of AI training might just hold the answer.