Sequoia Capital partner David Cahn estimates AI infrastructure spending will reach $1.5 trillion in 2026, necessitating $3 trillion in revenue to justify the investment. This projection is likely an underestimate due to rising memory costs and increased use of specialized inference chips. Hyperscalers like Google, Meta, Microsoft, and Amazon anticipate massive free cash flow accelerations by 2028, expecting payback from their chip investments.