The AI Boom’s Fuel: Advertising, and Existential Risk

OpenAI’s recent unveiling of its AI browser has ignited further speculation about an already overheated AI arms race, potentially accelerating the unprecedented levels of capital expenditure. The intertwined commitments of hundreds of billions in investment tied to future chip purchases have led some to question the sustainability of this frenzy, with some analysts even labeling it a bubble. A Harvard economist, for instance, estimates that a staggering 92% of US GDP growth in the first half of 2025 was directly attributable to investment in AI.

However, a deeper understanding is needed concerning the symbiotic relationship between this breakneck investment in AI and the fundamental business model underpinning the modern economy: the advertising technology (Ad Tech) industrial complex.

For the past quarter-century, the very fabric of the internet has been engineered to maximize advertising revenue. Search Engine Marketing, the bedrock of Google’s business model, remains arguably the most lucrative in history. Meta’s engagement- and attribution-based advertising business stands as a close second. And hot on their heels is Amazon’s advertising arm, leveraging its position as the world’s largest online retailer. While constituting a smaller fraction of Amazon’s overall revenue, its high-margin advertising business accounts for a disproportionate share of its profits. This success has spurred a wave of retail media networks launched by nearly every major retailer, significantly bolstering the bottom lines and market capitalization of giants like Walmart, Kroger, Uber (and UberEats), Doordash, and countless others.

These platforms have, in fact, been leveraging AI for years to fine-tune their advertising models. Algorithmic models power search and recommendation engines, striving to increase engagement and predict purchase decisions, ultimately seeking a larger slice of all commerce, not just what is conventionally considered “advertising.” These behemoths, commanding multi-trillion-dollar market caps, either entirely or substantially derive their profits from advertising. They are now channeling a portion of those historically lucrative advertising revenues into infrastructure investments on a scale not seen outside of wartime government spending.

The irony lies in the potential of this latest wave of AI to disrupt the very business model that fuels it. AI will undoubtedly revolutionize how people search (Google), shop (Amazon), and consume entertainment (Meta). Imagine answers delivered without navigating through multiple web pages, AI-assisted shopping experiences, and an endless stream of personalized content.

If AI poses such an existential threat to these established models, why are Google, Meta, and Amazon so heavily invested in the current AI gold rush? The “moonshot” scenario posits that achieving Artificial General Intelligence (AGI), or Super Intelligence – AI capable of exceeding human capabilities across the board – would unlock unprecedented value, dwarfing any current investment.

However, there exists a more pressing urgency to safeguard, or perhaps disrupt, the advertising-driven business model that underpins trillions in market capitalization and hundreds of billions in current investment before a competitor gains the upper hand. While the seminal paper that catalyzed this phase of AI, “Attention is All You Need,” was authored primarily by Google researchers, it was OpenAI and Microsoft, and now Grok, that ignited the current AI arms race. These companies are significantly less reliant on the existing advertising industrial complex. In fact, Sam Altman has characterized the content feeds generated by major platforms using AI to maximize advertising revenue as “the first at-scale misaligned AIs,” signaling OpenAI’s intent to disrupt these very businesses.

While the current surge in infrastructure investment shares similarities with the dot-com bubble of 2000, crucial differences exist. Unlike the nascent, often unprofitable companies of the dot-com era, today’s investors are some of the most profitable corporations globally. Thus far, there are no clear signs of cracks in the advertising-based business model that sustains both their investments and their market valuations (along with countless other companies not typically perceived as being in the “advertising business”).

However, if AI disrupts or even dismantles the current advertising-centric model, the resulting shock to the economy and markets would likely be far more profound than most currently anticipate.

Google, Meta, and Amazon remain strategically positioned to develop new business models, having successfully leveraged AI to bolster their advertising ventures for an extended period.

Fundamentally reshaping how individuals interact with search, commerce, and online content will necessitate entirely new revenue streams – potentially, and hopefully, models that prioritize alignment and move beyond advertising dependency. Regardless of the eventual model, it is perhaps prudent to recognize that the rationale behind AI infrastructure spending may extend beyond merely unlocking new revenue streams. A significant motivation could be to safeguard the underlying business models that constitute a far larger proportion of public companies’ market capitalization than is widely understood, a silent underpinning of the entire technology ecosystem.