In my previous article, I traced how the companies that built the internet lost the distinct cultures and identities that made them formidable, optimised into homogeneity by the same growth logic that made them successful. Now I will move to examine how AI has impacted this process.
The first thing to understand about Big Tech’s AI pivot is that it was not a strategic decision.
Across all AI-centric announcements in Big Tech, the language is identical: Transformative. At scale. Across our entire product surface. It’s not just the language; the investor calls, consumer messaging, acquisitions, and internal decks all follow the same patterns and ideas.
Every major technology company is now, officially, an AI company, having arrived at this conclusion within roughly the same window, having made roughly the same infrastructure bets, and having restructured roughly the same proportion of their workforces to fund them. The probability that this represents independent strategic convergence of all these organisations is frankly ridiculous. They are making strategic calls simply based on what their competitors are doing to maintain an illusion of “competitiveness”.
René Girard, the French anthropologist and literary theorist, called it mimetic desire: the observation that humans do not independently decide what to want, but want what they see others wanting. Applied to organisations, it produces what he called a mimetic crisis, which refers a state in which rivals, having imitated each other into near-identical positions, stop competing over what is actually valuable and start competing over the appearance of competition.
This is supported by DiMaggio and Powell, who observed that organisations under pressure from external stakeholders (like investors) don’t innovate, but imitate the competition, converging on whatever behaviour their peers have already legitimised, purely because it’s seen as safe and deals with the anxiety of innovation uncertainty.
The AI pivot is, therefore, a mimetic crisis at an industrial scale. The fear of missing a rival’s gain has become more powerful than any assessment of whether the gain is real.
The AI pivot is, by any honest reading, a mimetic crisis. Every major technology company is now, officially, an AI company. The announcements are indistinguishable, and all contain the same words in the same order: transformative, at scale, across our entire product surface. What is harder to find, in most cases, is evidence that this is producing value for the people who use the products.
Studies from Goldman Sachs and MIT have questioned whether enterprise AI spending generates productivity gains proportionate to the investment. The technology is being deployed ahead of demonstrated value, which is not a description of a confident strategy. It is a description of mimetic panic, as nobody wants to be the company that missed the wave.
Meanwhile, the jobs supposedly being augmented by AI are being eliminated by the same companies making the loudest noise about AI’s transformative potential. The mimetic crisis requires a resolution, and Girard argued that resolution historically comes through the scapegoat mechanism wherein the community unifies around the sacrifice of an arbitrary victim, temporarily relieving the tension of rivalry.
Also Read: A 65% probability explains the next likely move for Bitcoin as leverage clears
In corporate terms, it refers to the moment when the AI bubble deflates, and a single company is singled out as the cautionary tale while the larger players quietly reassess this investment.
What the bubble produces
To understand where mimetic capital flows when it chases a narrative ahead of validated products, consider Cluely.
Cluely was founded in 2025 by two Columbia University students, one of whom had been expelled for using an earlier version of the tool to cheat through an Amazon technical interview. The product was honest about this: an AI overlay that sits invisibly on your screen during video calls and feeds you answers in real time.
Within months, it had raised US$5.3 million in seed funding, then a US$15 million Series A led by Andreessen Horowitz, valuing the company at approximately US$120 million. All of this, for a company with an unclear business model, a viral marketing strategy, and revenue numbers that turned out, in March 2026, to be fabricated. The CEO admitted on X that the US$7 million ARR figure he had given to TechCrunch the previous summer was a lie.
The company has since rebranded as an AI meeting note-taker.
Cluely is not an anomaly, but is a case study in what DiMaggio and Powell’s mimetic isomorphism looks like at the level of venture capital: under conditions of uncertainty, investors copy other investors. As the field signals that this is the kind of company worth investing in, the logic of missing out becomes more powerful than the logic of performing due diligence. By 2025, AI captured close to 50 per cent of all global venture funding: over US$200 billion, up from 34 per cent the year before.
The terror of being the one who missed the next OpenAI has become more powerful than any assessment of individual companies. The object of desire ( AI value) has receded and has been replaced by an obsession to win the investment war.
The most significant threats to the monoculture are being built by companies quietly constructing alternatives to the platforms the monoculture depends on.
Linux desktop usage has historically been a running joke: it’s always the year of Linux on the desktop, but never quite. Framework, a new age PC and laptop manufacturer with a distinct cultural identity around repairability, now ships laptops with Linux as a first-class option, which sold out prior to its Windows counterpart.
Lenovo ThinkPads come with Linux pre-installed. The developer community has shifted significantly toward Linux-based workflows. This is not yet a mass-market story, but it represents, in DiMaggio and Powell’s terms, coercive isomorphic pressure running in reverse: an external force that makes the dominant platform’s continued stagnation costly, rather than safe.
The gaming front is more significant. Windows has historically been the stickiest consumer use case: the reason people tolerated everything else about it. Valve’s Steam Deck changed that. A Linux-based device that runs an enormous proportion of the Steam library with fewer power requirements.
SteamOS is now expanding beyond the Deck due to its contributions to the WINE platform and the Proton compatibility layer (both of which are free). Valve quietly built a gaming ecosystem that no longer requires Windows to function. That is a disruption in the structural sense, which has led to Windows and Xbox’s newfound focus on improving their services as of late.
Also Read: Emotional intelligence makes AI training stick
The reckoning
The current market is an example of DiMaggio and Powell’s iron cage: organisations that have become so optimised for legitimacy to appear correct to investors, regulators, and each other, that they have lost the ability to actually be anything in particular.
The companies that built the internet had something the current iteration has optimised away: cultural density, the very thing that makes a company relatable and marketable to consumers in choice-based markets like consumer technology.
Girard was clear about how mimetic crises resolve: scapegoats. The dot-com correction produced its scapegoats in companies like Pets.com and Webvan. The scapegoat is never the system itself, but an organisation within the system, selected to make the system appear self-correcting. Cluely may serve this function, or it could be a tech giant, or another one of the thousands of new-age tech companies.
What breaks the cycle is the alternative. Valve didn’t announce a strategy to challenge Windows; it simply built a device, then a platform, then an ecosystem, until the alternative structurally competes with the market leader. Framework didn’t publish a manifesto; it simply made excellent products in line with its simple mission of repairability and customisability.
The irony is that isomorphism has made differentiation easier, not harder. When every large tech company looks, sounds, and feels the same, being genuinely excellent at something specific and being honest about what you believe in is a competitive advantage of unusual power. The only problem is that the conditions that produce that kind of strangeness are precisely what institutional success destroys.
—
Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.
The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.
Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.
The post The AI bubble and its reckoning — Part 2 appeared first on e27.