The Download: The trillion-dollar AI bet and OpenAI's push for biology data

Meta description: New analysis shows AI hyperscalers' earnings must grow extraordinarily fast to justify nearly $1.1 trillion in spending by 2027.

Tags: AI investment, hyperscalers, AI data centers, OpenAI Foundation, tech economy

Featured image: abstract technology style — glowing server racks and data streams, no people or logos

The staggering math behind AI's data center boom

As frontier AI systems grow more powerful, the world's largest tech firms are pouring unprecedented sums into the infrastructure behind them. But whether that outlay will ever generate returns is one of the biggest open questions in the global economy — and new research suggests the bar for success is far higher than most people realize.

When Jessica Wachter, a finance professor at the University of Pennsylvania, set out to measure AI's likely economic impact, she confronted a wall of unknowns. So she began with what she calls a "remarkable fact" beyond dispute: a small group of hyperscalers is committing enormous capital to AI data centers.

How much growth does the spending require?

Rather than guessing how widely AI models will be adopted, Wachter flipped the question: how quickly would the hyperscalers' earnings need to climb to make their investments worthwhile through 2027 — the year expenditures are projected to approach $1.1 trillion?

The answer is striking. To break even by 2030, AI companies would need to deliver an extraordinary surge in productivity — a pace of growth that few industries in history have achieved.

The analysis underscores the scale of the gamble now underway: the buildout isn't just expensive, it demands near-unprecedented economic gains simply to cover its costs.

OpenAI Foundation funds "biotech's lost archive"

In other AI news, the OpenAI Foundation — the nonprofit parent of OpenAI — announced this week it will fund an idea floated by policy analyst Ruxandra Teslo: acquiring data from failed biotech companies. By bidding at bankruptcy proceedings, her proposal argued, it could be possible to secure detailed regulatory filings, manufacturing strategies and safety data, forming what she called "biotech's lost archive." The foundation will pay to create "high-quality scientific datasets" from such material, betting that richer biological data is key to AI breakthroughs in curing disease.

What to watch next

Whether hyperscaler revenue can actually grow fast enough to validate the investment remains the defining question for the sector — and for markets watching AI's trillion-dollar bet unfold.