Why it matters
If the founder of a 1.5B-ARR, 10B-valued startup models token spend as a line that overtakes payroll, agent compute stops being an experiment budget and becomes a core operating cost CFOs must forecast.
Tokenmaxxing read
The bull case for tokenmaxxing, from a 1.5B-ARR founder: Foody argues the application layer has no moat and the model is the product, so the rational move is routing ever more work to agents until compute, not staff, is your biggest bill.
Source takeaway
Useful as a primary-source, founder-stated forecast, but it is a self-interested call from a vendor whose growth depends on rising AI adoption. Weigh the five-year token-over-payroll prediction as a thesis, not data.
