Key Takeaways

  • Alexander Altmann notes the entire S&P 500 produces roughly $17 trillion in top-line revenue and $5 trillion to $5.5 trillion in net profit, putting a hard ceiling on enterprise software budgets.
  • Cumulative revenue for Apple and Microsoft combined since 1990 totals under $10 trillion inflation-adjusted, exposing multi-trillion-dollar annual AI market projections as mathematically detached from historical tech spending.
  • Third-party tracking from Ramp and OpenRouter reveals corporate token expenditure has started to level off, despite ongoing top-line gains at frontier labs like Anthropic.
  • Enterprise budgets cannot support current multi-trillion-dollar AI valuations alone; Altmann argues that consumer-facing applications must carry the weight.

Enterprise Budgets Have a Hard Math Problem

Altmann points out the basic arithmetic that market projections routinely ignore. The entire S&P 500 brings in roughly $17 trillion in top-line revenue. After $11 trillion to $12 trillion in expenses, total net profit sits between $5 trillion and $5.5 trillion. For venture-backed AI startups and public cloud vendors to capture trillions in annual enterprise revenue, corporate America would have to redirect an impossible share of its total profit into model subscriptions and computing infrastructure.

Historical cycles put those projections in perspective. “Looking at the cumulative revenue of the two most successful tech companies in modern history, it was Microsoft and Apple,” Altmann says, “and it was the cumulative TAM of those two was, I think, inflation-adjusted, not even $10 trillion since 1990.” Expecting enterprise AI software to clear multi-trillion-dollar totals within a few years assumes business software spending will outpace the thirty-year footprint of personal computing, smartphones, and the enterprise cloud combined. As Alloway observes, “billions in earnings versus valuations at trillions, like, I think that is the gap that people are debating.”

Token Demand Is Leveling Off on the Ground

While frontier labs like Anthropic report strong top-line momentum, third-party transaction data tells a quieter story. Corporate experimentation has reached an early plateau. Software teams and IT departments ran pilots, acquired API keys, and tested internal copilots over the past two years. Now the invoices are arriving, and finance teams want clear return on investment before signing expanded renewals.

Altmann points directly to transaction data tracking developer spend: “Last time I checked, like, we all look at the Ramp data, or the OpenRouter data, and we can clearly see that that token expenditure is kind of begun to level off a bit.” When corporate spending slows, infrastructure providers face immediate pressure. Neocloud financing bottlenecks are appearing across the market, while the rise of local open-weight models gives companies an alternative to expensive cloud contracts. Altmann argues enterprise software cannot bridge the revenue divide on its own: “the consumer is everything. It's the economy. It's where, to the point about revenues, like enterprise revenues, I just don't think it's going to plug that gap.”

What to Do With This

Audit your software pipeline by checking your last six months of closed enterprise deals. If your product bills on API tokens or employee seats rather than a quantifiable cost reduction, assume your buyer will freeze or cut the contract on renewal. Repackage your pricing around direct, documented expense reductions, or look at consumer-facing distribution channels where individual willingness to pay does not require an enterprise procurement committee.