Quote · The a16z Show
Is AI a Bubble? | Gavin Baker on Data Centers, GPUs, and the AI Economy
Where this was said
Frontier Model Economics: Why AI Margins Will Never Look Like SaaS
At 14:43 · chapter starts 12:50
The economics of frontier AI models are fundamentally different from the software businesses that preceded them. Baker points to the SaaS playbook of 2021–2022 — companies routinely running at 80–90% gross margins — and explains why that benchmark is structurally unachievable for AI labs. The culprit is compute intensity: scaling laws and the growing importance of test-time compute mean AI models are inherently more expensive to run. Richard Sutton's 'Bitter Lesson' — the observation that methods leveraging raw computation consistently outperform those encoding human knowledge — is a structural anchor for these costs. The silver lining, Baker notes, is that lower gross margins don't preclude great businesses: if opex can be kept low, the math can still work. But investors and founders expecting SaaS-era margins from frontier AI are looking at the wrong benchmark entirely.
Pre-AI SaaS companies routinely achieved 80–90% gross margins, a level Baker says frontier AI companies structurally cannot match due to compute intensity.
Baker once predicted all application SaaS could go to zero. He's walked that back, but the urgency remains: Cursor has accumulated 1 trillion coding tokens, and incumbent public coding companies haven't even tried to compete. The window to lean in and run AI products at breakeven is closing fast.
It's definitionally impossible to succeed in AI without gross margin pressure. Baker urges SaaS companies to treat declining margins as a badge of honor — just as Microsoft successfully transitioned from on-premise perpetual licenses to a lower-margin cloud model and delivered a decade of stock gains.
