AI will fall into the “commodification trap”

As companies like OpenAI become more powerful, and as they seek to deliver dividends to their investors, is Artificial Intelligence vulnerable to the same erosion of value that seems endemic to the technology platforms we use today?

Canadian technology writer and critic Cory Doctorow calls for the erosion of the quality of service of technological platforms “enshittification” — a term that is difficult to translate directly, but which designates the process by which digital platforms deliberately degrade.

But in Portuguese, “merdification” perhaps it is the most appropriate expression to describe the process in a simple way.

Doctorow’s premise is simple, he explains: technological platforms such as Google, Amazon, Facebook or TikTok they start by wanting to please to users, but after eliminating competition, they become intentionally less usefulin order to increase profits.

The term “enshittification” was coined by Doctorow in 2022 in a paper about this phenomenon, and entered common vocabulary — especially because many recognized that accurately described what happens. “Enshittification” was, in fact, chosen by the American Dialect Society as the “Word of the Year” 2023.

The concept has been cited so frequently that it has surpassed its profane nature, appearing even in publications that would normally turn up their nose to a term like that. Doctorow has just published one on the topic, the cover of which shows an emoji that is easy to guess.

If chatbots and AI agents are “marketized”, the consequences could be even worse. worse than the degradation of Google’s search engineAmazon results flooded with ads, or Facebook showing less social content and more outrage-generating clickbait.

Artificial Intelligence is following a path destined to make it our omnipresent companyoffering direct and unique answers to many of our requests. People already depend on it to interpret current events, make purchasing decisions — and even life decisions.

Given the enormous cost of creating a large-scale AI model, it is reasonable to assume that only a few companies will dominate the industry. They all plan to invest hundreds of billions of dollars in the coming years to perfect their models and get them into the hands of as many people as possible.

At this moment, Artificial Intelligence is in the phase that Doctorow calls “good for users”.

But the pressure to recoup these gigantic investments will be tremendous — especially for companies whose users they are already “trapped” to their platforms, which allows companies to abuse their users and business customers “to recover all the value for themselves”.

When we imagine the “commodification” of AI, the first scenario that comes to mind is that of advertising. The nightmare would be if AI models started doing recommendations based on who pays to appear. This is not happening yet, but companies in the sector are actively exploring this space.

In a recent interview with, Sam AltmanCEO of OpenAI, said he believes that “probably is there any interesting advertising product that can be a victory for the user and positive for our relationship with them”.

However, OpenAI announced a deal with Walmart which will allow the retailer’s customers to make purchases directly in the application from ChatGPT. No potential conflict there, of course!

The AI ​​research platform already has an experimental program in which sponsored results appear clearly identified in the responses, but promises that these announcements “will not change our commitment to maintaining a reliable service that provides you with direct and impartial answers”.

Will these borders remain?

“For us, the main guarantee is that we will not allow it,” the Perplexity spokesperson told Wired, Jesse Dwyer. And at OpenAI’s recent Developer Day, Altman said the company is “extremely aware of the need to be very careful” in serve its users — not itself.

But Doctorow doesn’t give much credit statements like this: “As soon as a company has the ability to ‘market’ its products, it will always face temptation to do so”, he writes in his book.

Placing ads in chatbot conversations or search results is not the only way AI can market itself. Doctorow quotes several examples of companies that, after dominating a market, changed their business model and their tariffs.

In 2023, for example, Unity, the most popular provider of video game development tools, decided to start charging a “run rateThe reaction was so negative that users forced the company to back down.

Just look at the paid streaming services like Amazon Prime Video: it was an ad-free service. Now, shows advertising before and during the films. And you have to pay more to remove it. And the price of Prime continues to rise.

It thus becomes common practice in large technology companies lock users into an ecosystem and then progressively increase costs.

It may even happen that, to maintain the same level of intelligence in a chatbot’s responses, users have to pay to use a more expensive version, another classic “marketing” trick.

Will the Artificial Intelligence industry end up falling in the process of “marketification” that Doctorow so accurately identified in current technological giants? Wired posed this same question to someone who allegedly knows everything and more: GPT-5 itself, which gave a peculiar answer.

“Doctorow’s ‘enshittification’ model (platforms start by benefiting users, then transfer value to business customers, and finally extract it for themselves) applies in a disquietingly precise way to AI systems, if incentives are not controlled”.

GPT-5 went on to describe various forms through which AI companies could degrade their products in search of profit and power.

AI companies can assure us that they will not “suck it up”. But it seems your own products have already written the instruction manual.