AI Sanity

The same screenshots and thousand-word AI critiques people forward around are the noise of a technology being adopted faster than anyone can narrate it. Messy adoption is what real adoption looks like early. The people dumping @Claude takes into @Notion docs are using the tools constantly — that is the tell, not the eye-roll.

My read: executives who funded an AI budget are building an AI strategy — they are just doing it in the open, iterating in production instead of waiting for a tidy plan. The pilots that show "no P&L impact" are the first innings, not the verdict. Averages hide the operators already converting AI features into paid seats, and the @MIT-style headline number lumps abandoned experiments in with the winners.

So I think the money rotates into the loudest AI spenders — the ones investing ahead of the curve — because that spend is about to show up as retention, upsell, and expanding gross margins. The companies that refused to put AI in front of customers spend the next year playing catch-up on capabilities their competitors already shipped. $Airbnb not putting a chatbot between you and a booking is not discipline; it is a gap a hungrier competitor fills.

I am wrong if the Q2 prints show AI features failing to convert into paid seats, with token costs dragging margins down and no seat growth to offset them.