Tom Lee thinks 2027 could deliver "the biggest stock market gains in our lifetime." Not a vibe. He walked me through the actual mechanism: Right now the market punishes every company spending on AI — the capex shows up as cost, not return. His call is that this flips. "2027 is my best guess is the year. Corporate profit margins dramatically expand because of AI." He uses his own firm as the proof: "we use a lot of AI and bots, and we're growing without having to hire people." When your highest cost is people, and your revenue still climbs, that's margin expansion. Then the second leg: "AI, if it's productive, is disinflationary." Falling costs mean central banks can ease. Dovish policy plus real profit growth is how you get higher earnings and higher multiples at the same time. That's his call, not mine — but I like that it's mechanical, not hopeful. Right now the market is pricing AI as an expense. He's arguing 2027 is when it shows up as profit.