For centuries price has been a sign of quality. Generative AI is breaking that rule in dozens of sectors

For centuries, price has served as a cognitive shortcut. If something costs a lot it is because, for one reason or another, it must be worth a lot. An Armani suit, Bang & Olufsen headphonesa McKinsey report. The number has always served to convey certain information to us before seeing the product. It was compressed reputation.

With the arrival of generative AI, that is ending in many sectors.

Today a logo can cost 15 euros or 15,000. And be the same logo. A market analysis can come from a consulting firm with offices on three continents or from a guy in pajamas who knows how to wear Deep Research. The report may be indistinguishable. In fact sometimes the second one will be betterbecause the guy in pajamas understands the sector and the consultant assigned the junior who was free.

AI is breaking the link between production cost and final result. Something very similar to what Antonio Ortiz, AI popularizer and former final boss of this house, in “Artificial Intelligence and unlinking effort and result“.

If anyone can generate in minutes what previously required teams, weeks, and invoices with many zeros, price no longer communicates much about quality. It’s starting to be noise.

and this will force a signal migration. From ‘how much’ to ‘who’, to ‘how’ or ‘why’. The questions that will matter are going to be “who signed this?”, “what process followed?”, “what human decisions were behind it?” That is, the process will become the product.

It is already beginning to be seen with design studios that They obsessively document any iterationor consultancies that not only sell you the deliverable but also also access to the reasoning of their partners. More and more we are digital artisans who charge for showing how we work and not only for what we deliver.

AI has made production almost free, so we are being flooded with digital content of all kinds, so scarcity shifts to criteria.

Knowing what to ask for, what to discard, what makes sense and what doesn’t. to good taste. AI can do almost anything, and what it can’t, it will learn next year. Deciding well what to do and what not to do is still expensive.

There, for the moment and luckily, there is no shortcut.

Featured image | Xataka

In Xataka | The AI ​​of 2026 brings an uncomfortable truth: the most useful will be the one that watches us the most