No Ads My A*s (or the Slow Googlization of OpenAI)

Remember what Google promised?

Google is a pure search engine, no weather, no news feed, no links to sponsors, no ads, no distractions, no portal litter, nothing but a fast-loading search site.

A vow of ascetic, almost zen-ish minimalism, a manifesto of purity flung into the turbulent waters of the Y2K millennium bug hysteria that electrified the year of the Lord 1999. And yes, the sentence that now reads like an archaic hymn to innocence ended with the most ironic of claims ("no ads") as if the business model of the entire web would not later crystallise around that very fault line. This irony becomes even more striking when one recalls that, in the latest publicly available data, advertising accounted for approximately 76 percent of Alphabet Inc's colossal revenue figure, a reminder that the corporate path from monk-like restraint to maximalist monetisation tends to be very short and very slippery. But I digress...

We stand in front of a similar theatre today. However, the architecture has grown to the size of Comet 3I/ATLAS. The actors who inhabit it wield a form of cognitive power that the early search engines could not even dream of (ahhhh, remember the piratesque era of AltaVista, a time when the apex of technical ingenuity consisted of sprinkling p*rn keywords in white text on a white background and praying that the crawler, bless its innocent silicon soul, would glide past without realising it had just been digitally mugged?)

Yet the chorus is the same, almost like a reboot of an '80s classic nobody asked for, the cinematic equivalent of waking up to find that Hollywood has decided the world urgently needs an updated Back to the Future trilogy (spoiler: we DON'T!).

Sam Altman, the pretty-fly-for-a-white-guy man who repeatedly said in public interviews that he hates advertising, man, now assures us that generative intelligence will keep the answer sacred, that monetisation will remain peripheral, and that truth will not be bent by sponsorships. In a recent interview, he even went as far as declaring that ads on a Google search are dependent on Google doing badly, since if Google gave the best possible answer, there'd be no reason ever to buy an ad above it. A beautiful sentiment, as long as one never actually looks under the carpet…

Because reality speaks with a different timbre. Google Search, YouTube advertising, and the display network generated roughly 265 billion dollars in revenue last year, a figure that essentially mirrors the annual economic output of freaking Finland! If delivering imperfect answers were the condition for sustaining that empire, the implication would be unsettling.

A What-If?-World in which Google suddenly did not do it badly would be a world in which one of the largest revenue machines on the planet evaporated overnight, taking with it an economy the size of a sovereign nation (and by heavy-metal extension, living in a world without Finland would mean living in a world without black metal pioneer band Impaled Nazarene, and I do not want to live in that world. Do you?)

Altman also insisted that ChatGPT would not accept money to put a worse hotel above a better hotel, describing such a scenario as catastrophic for your relationship with ChatGPT. He even claimed that he has no idea what ads will look like and that ads are not our biggest revenue opportunity, a sentence that belongs to the same literary genre as I am only having one pint of Guinness tonight (I love that book!).

It is difficult not to feel the echo of Mountain View's early vows. Mark Twain once wrote that history rarely repeats itself, but it often rhymes, and the rhyme here is unmistakable. That being said, the announcement that ChatGPT will include advertising should not have surprised anyone. Intelligence at planetary scale is not a Platonic ideal floating above matter. It is computation, and computation is a material process that consumes (too much?) energy, accelerates costs, and demands an economic substrate robust enough to sustain its own gravity. The dream of a pure and uncorrupted, almost Marxist AI, generously subsidised by benevolent capital, is as fragile as Google's early purity. Because purity remains intact only as long as the infrastructure remains light. But as soon as the system acquires weight, monetisation follows with the inevitability of a physical law, and as I write this, ChatGPT has reportedly over 800 million weekly active users, which is the kind of scale that makes purity feel like a nostalgic hobby.

So, how will these ads look anyway?

Hard to say, but we are already watching the hyperlink economy dissolve into something else entirely. What once was a centrifugal architecture built on exits, citations, and blue links pointing toward what is now defined as "the open web" is mutating into a centripetal chamber where discovery no longer happens across a SERP but inside a single linguistic organism. AI as the ultimate UI. Google's AI Overviews have already revealed the blueprint: AI ads are no longer positioned beside the answer; they begin to bleed INTO the answer, merging with the synthetic prose of the model, becoming indistinguishable from the epistemic tissue they inhabit, a kind of semantic osmosis where e-commerce and e-cognition share the same bloodstream.

This is the logic behind what I described in (too many?) articles as "pay-per-mention", a clear departure from the familiar pay-to-rank, CPM, and CPC advertising models we know, a shift toward an economic architecture that I suspect will dominate this "post-search" decade.

The reassurance Altman offers, the promise that generative systems will remain incorruptible, reproduces the exact optimism that shaped Google's original pledge. With the aggravating factor that I genuinely think Google, in its proto-adolescent, almost Edipian phase, might have actually believed its own rhetoric, while Altman is performing a sort of strategic innocence, a curated naivety designed to make him look like the kid who stumbled into the cockpit by mistake (hey, after all this is the same guy who, speaking on Capitol Hill, declared that he does not hold equity in OpenAI and that he earns simply enough to cover his insurance expenses...)

I do not want to sound strongly anticapitalist here, although the sentence writes itself, but no system that earns a commission on hotel bookings can remain epistemically neutral. Even if the commission is flat, the medium shapes behaviour long before money even circulates.

So my take is that what will emerge is not a new ad format but rather a new epistemology. The interface becomes the marketplace, and the marketplace becomes invisible.

For hospitality, this is not an incremental shift. It is a rupture that cuts through the entire ecosystem of discovery. Hotels will no longer compete for pages or rankings; instead, they will compete for the right to be mentioned by the models. Mention-as-currency. OTAs, brand.com and metasearch engines will recede into a background substrate that feeds ARI to an agent that never reveals its sources. Brand protection, as we know it, might continue, although the upper funnel melts into a precognitive zone where desire is shaped before the user even realizes that they are searching.

Generative advertising will resemble less campaigns as we know them and more something new: some kind of narrative engineering. Product, language, and monetisation fuse into a single autonomous current. The model becomes strategist, storyteller, and transactional. Because, as I wrote many times, the coming decade belongs to ecosystems in which discovery, persuasion, and transaction are not stages in a funnel but different crystallisations of the same linguistic field.

To cut a long story short: returning to that 1999 claim is not an exercise in nostalgia. It is a reminder that purity is always temporary. Google once promised a search page free of ads. OpenAI today tells us it will preserve integrity. Perhaps it will attempt to do so. Perhaps it even believes the promise. Yet the gravitational field of monetisation bends everything within its radius, and generative systems are not above ideology. They are the place where ideology becomes syntax.

The question is whether we have learned anything from the first cycle of the web, whether we can recognise the precise moment when purity begins to corrode, or whether we will once again accept the erosion as a naturalised truth of the digital universe.

Or if we are brave enough to bite, as Trent Reznor would put it, the hand that feeds...