Category

Your Supermarket Knows Those Strawberries Are Dying

AI could help supermarkets cut food waste by reducing prices as products approach expiry. But the same technology can also personalise what people pay. The real question is where retailers draw the line.

3 min read

Digital supermarket price tag showing the tension between AI markdowns for food waste and personalised surveillance pricing.
Digital supermarket price tag showing the tension between AI markdowns for food waste and personalised surveillance pricing.

The price tag is becoming software.

For most of our lives, supermarket pricing has been reassuringly stupid.

Someone decides a price. Someone prints a label. Someone sticks it on a shelf.

If the strawberries are still sitting there looking increasingly tragic at 7pm, someone eventually puts a yellow sticker on them.

AI changes that.

Digital shelf labels, live inventory data and pricing algorithms mean supermarkets can increasingly know what is selling, what isn't, what's approaching expiry and when a price cut might save something from ending up in the bin.

On the face of it, that's pretty sensible.

Walmart is already rolling out digital shelf labels, although it has been very careful to point out that prices aren't being changed according to the weather, time of day or who happens to be standing in front of the shelf.

Which is interesting.

Because nobody gets that specific unless they know exactly what people are worried about.

The same technology has a much creepier use

Once the price tag becomes software, there's another possibility.

Instead of asking:

“How much should we reduce these strawberries so somebody buys them?”

the system can start asking:

“How much is Pablo prepared to pay for these strawberries?”

That's a very different proposition.

Regulators have noticed too. The US Federal Trade Commission has been looking at personalised pricing and the use of personal data to determine what people pay.

And this is where something potentially quite useful for sustainability gets complicated.

AI could actually help us waste less food

Food waste is partly a pricing problem.

A yoghurt worth CHF 2 this morning probably isn't worth CHF 2 when it expires tonight.

But supermarkets have historically been pretty clumsy at dealing with that.

AI doesn't have to be.

It can combine inventory, demand and shelf-life data and start reducing prices before food becomes waste.

Check it. 👇

That's potentially a brilliant use of AI.

The supermarket protects some margin.

You get cheaper food.

And fewer strawberries die unnecessarily.

Everybody wins.

Until the exact same infrastructure starts deciding that *you* can probably afford to pay more.

And that's where the whole thing falls apart

There are really two different ideas hiding under the phrase “dynamic pricing”.

One is based on the product.

It's nearly expired. There's too much stock. Demand is low. Reduce the price.

The other is based on you.

Where you live. What you've bought before. Your loyalty data. How urgently you appear to want something. Perhaps even what the algorithm thinks you're willing to pay.

Those things might use similar technology.

But they feel completely different.

If AI pricing is sold as sustainability while customers experience it as surveillance, the environmental story won't save it. It'll just make the whole thing feel more cynical.

And I think there's a wider lesson here.

Companies are going to discover lots of genuinely useful applications for AI.

But just because the technology *can* optimise something doesn't mean people will accept every version of that optimisation.

Sometimes the smartest thing you can do with AI is put a very clear boundary around what you're not going to use it for.

So if supermarkets want to use AI to rescue those strawberries, great.

Tell us when prices move.

Tell us why.

Tell us what data isn't being used.

And most importantly, make sure the cheaper strawberries cost the same whether I'm buying them or a millionaire is standing next to me.

© 2026 Pablo Retamal. Geneva, Switzerland. All rights reserved.

© 2026 Pablo Retamal. Geneva, Switzerland. All rights reserved.

© 2026 Pablo Retamal. Geneva, Switzerland. All rights reserved.