The CFO Just Left the Room

Scott Galloway said something worth sitting with on Pivot this week.

The CFO of Uber burned through their entire AI budget in eight weeks. No measurable improvement in the product. Not a bad vendor. Not bad technology. Bad question.

The question was "Can AI work?" The question they needed was "Can AI pay?"

Galloway's read: we've gone from a supply crisis to a demand crisis in roughly 30 days. The largest cloud companies will spend $600 to $700 billion on AI infrastructure this year. Enterprise customers cannot show their CFOs what they got for it. That gap does not close on its own.

He compared it to 1999. Not as an insult. As a pattern. Back then, nobody could pick the winning application, so everyone invested in infrastructure instead. Cisco hit an all-time high. Then lost 93% of its value. The technology survived. The valuations didn't.

What We See in the Field

The problem is not access to AI. It is not even cost. It is that most organizations started with tools before they had clarity about outcomes. When the CFO asks for ROI, there is no answer because no one wrote down what success looked like before the contract was signed.

Getting started with AI is easy. Organizational change is what is hard.

The companies that will navigate this moment are not the ones that spent the most. They are the ones who started with the right questions. Who does this serve? What does measurable improvement look like here? How do we know when it is working?

That is not skepticism about AI. That is how any real investment gets made.

We are watching smart leaders ask those questions right now. That conversation starts at coastalintelligence.ai.

---

Published by Coastal Intelligence. AI Think Tank and Consultancy. Santa Barbara, California.