AI Boom Starves Everything Else, But Hey, It's 'Inevitable'

Axios has made a discovery so profound it required investigation: when trillions of dollars flow into one sector, other sectors receive less money. This crowding-out effect—the economic principle that capital devoted to AI infrastructure and model development necessarily displaces investment in literally every other industry—has finally been quantified and presented to readers as though it were breaking news rather than a basic tenet of finance. The real news, apparently, is not that this is happening, but that the crowding-out effect is "smaller than you might expect." Smaller. Than expected. One can almost hear the sigh of relief echoing through venture offices.

What makes this framing so deliciously absurd is the underlying admission: yes, we are consciously starving healthcare innovation, climate tech, biotech, and traditional infrastructure of capital. Yes, we are doing this at scale. Yes, we understand this means those sectors will receive fewer resources than they otherwise would. But the story arc demands a redemptive third act, so we are treated to the assurance that the damage is "smaller than you might expect." Smaller compared to what, exactly? Compared to the apocalyptic scenario where literally zero dollars flow anywhere but GPUs? The bar has been set so low that merely catastrophic misallocation now counts as a relative win.

The structural dishonesty here deserves recognition. An economist could spend five minutes explaining that "crowding out" is not a quantifiable phenomenon with a precise threshold—it is a directional reality. Every dollar chasing transformer weights is a dollar not funding an early-stage oncology platform or a geothermal startup or literally any other innovation that cannot generate a $10 billion IPO narrative within 18 months. The piece frames this as inevitable, which is technically true in the sense that gravity is inevitable. It is also, as gravity, something we might choose to resist rather than simply accept as natural law.

What is most revealing is what the story does not examine: whether this crowding out represents efficient capital allocation or catastrophic misallocation. No one seriously argues that data center infrastructure and large language model development represent 100% of all productive investment opportunities in North America. Yet we have constructed a financial system where they are capturing the overwhelming majority of venture and institutional capital anyway. The fact that Axios frames this as "smaller than expected" rather than "inexplicably massive" tells you everything about the current epistemic collapse in financial journalism.

The implicit assumption undergirding this entire piece is that the AI boom is a permanent feature of the landscape, not a cyclical phenomenon subject to valuation compression and disappointment like every other technology bubble before it. Should the market correct—should we discover that $500 billion in annual infrastructure spending on large language models represents, in retrospect, a catastrophic misallocation—the crowding-out will be remembered as catastrophic, not inevitable. The companies that could not raise Series B rounds because all the money went to compute will look like the collateral damage they actually were.

What does this say about the current state of venture capital? Precisely this: we have sophisticated enough frameworks to quantify crowding-out effects, but we lack the moral courage to ask whether we should stop creating them in the first place. The story is technically competent, analytically sound, and fundamentally evasive. It is the perfect artifact of an industry that has decided to worship at the altar of one technology while pretending to wonder why nothing else can get funded.

Inevitability, it turns out, is just another word for "we decided not to decide differently."