This isn't AI, this is the minimum wage

Yes, OK, slow time of year, PR attempts to get a company written up will work well right now:

AI sparks revolution in how much supermarkets charge you for food

Digital labelling can pass on price fluctuations more quickly and cut waste

Yes, super and that might even be the effect. But the driver is, we very strongly suspect, this:

electronic shelf labels (ESLs), the first step towards dynamic pricing, had been introduced in a small number of existing shops,……..Retail expert Clare Bailey said the move to digital labelling is the first step towards a “dynamic pricing model” in supermarkets as they look to reduce labour, ….. costs

We’ve all seen this in a shop, an individual clacking out new price labels to stick on a shelf of something or other. That person costs the shop about £10 an hour this year, it’ll soon be about £11 just in wages alone (then add NI, benefits etc) and so the hunt is on to reduce labour costs. One central data entry to change the price on a whole shelf of product, job’s a good ‘un.

We can think of this as just tech reducing labour costs, or labour costs being forced up and thereby inducing the job killing use of technology. Either works.

Far more fun, to us at least, is that this begins to destroy a core tenet of New Keynesian economics. The idea of menu costs is central - it explains price stickiness. Prices do not smoothly change, they move in shuddery jumps. For there’s a cost to changing the price - the cost of reprinting the menu - so it’s only worth doing that when the underlying has changed enough to justify that cost of actually making the price change.

This is true too. It’s also why the New Classical and Real Business Cycle theories seem not to explain the world quite right, but New Keynesian seem to do better. Which is why every central bank and Treasury economic model is, by and large, New Keynesian.

That’s all a recent development, certainly recent decades. Which is where the fun comes in really. For those real world economic models have all zeroed in on a specific explanatory structure just as that structure becomes non-explanatory. If changing prices is now the one single entry in a central database for a shop, possibly for an entire chain, then menus costs are much less of an issue. The world is moving closer to those RBC and New Classical models where prices change swiftly and near costlessly and so the economy as a whole reacts near instantly to change.

Yes, obviously, these are all tendencies, not absolutes. But we really do think that it’s terribly fun that just as the orthodoxy narrows in on something generally agreed that it is also becoming untrue.

As in that old joke about economics exams. Universities still use the same questions they did a century ago - it’s the answers that have changed. Which, actually, they have.