Microsoft CEO Satya Nadella stated on the Hard Fork podcast that most enterprise AI adoption has focused on volume—such as how many tokens are consumed or how many engineers use Copilot—rather than business value. He said, “The marginal cost of productivity improvement has to match the marginal cost of the token,” calling it a management discipline. Nadella noted that token costs compound fast, citing Uber burning through its entire 2026 AI coding budget in roughly four months, with per-engineer monthly API costs between $500 and $2,000. He admitted Microsoft itself has engaged in “token maxing,” responding “A lot” when asked about internal overconsumption. In February 2025, Nadella set his benchmark for AGI at 10% GDP growth, later describing the difficulty of translating general-purpose technology into measurable output due to required change management. His formula for achieving 10% growth is “a perfect match between the marginal cost of the token to the marginal value and it’s priced right.” He emphasized that token spending without discipline is just cost, stating, “Everybody goes and vibe codes and token maxes. That’s not a way to achieve 10% growth.”