The AI Economy Has a Commitment Problem, and the Bill Is Coming Due

Every CEO in America is out here swiping right on AI. Nobody’s ready to get married.

That’s the read from Amy Webb, the futurist who runs Future Today Strategy Group and talks to somewhere between 100 and 150 CEOs a year for a living. She dropped a line to Fortune that stopped me cold: the corporate AI boom looks a lot like dating-app fatigue. Endless matches, endless pilots, nobody closing. She called it “pilot purgatory,” and if you’ve sat in a single enterprise strategy meeting in the last eighteen months, you already know exactly what she’s talking about.

Here’s the setup. Companies are running pilot after pilot. Fourteen, fifteen of them in a single year for one of Webb’s clients. Small teams, the Amazon two-pizza rule, quick experiments meant to move fast and prove value. None of them scaled. Why? Because nobody’s plugging these pilots into legal, into IT, into the actual infrastructure that makes something real. Every pilot restarts from zero. Every restart costs money. And the pizza just keeps coming.

Webb put it plainly: it feels like you’re buying abundance, but that abundance shows up on the invoice later. Not as a slow drip either. As a bill you didn’t see coming.

And it’s not just the pilots. It’s the decks. Executives are drowning in them. Presentations that used to take a week now take a day, except teams are getting five times as many of them dumped on their desks. Webb’s word for it was “insta-decks,” and she pointed out something that should make every AI vendor squirm a little: these tools have a verbosity problem. You ask for one page and get ten. More output, less signal. More noise, less decision.

That’s the part people keep missing when they talk about “AI slop.” This isn’t sloppy. It’s polished, professional, and completely generic. Webb’s line was sharp: the more a company leans on these tools, the more generic the ideas that come out the other end. Nobody’s thinking harder. Everybody’s just moving faster toward the same conclusion everyone else already reached.

So when does the reckoning hit? Webb told Fortune she expects it as early as next year, calling it the second chapter of this story. Wall Street starts asking hard questions about all those pilots. Missed targets start showing up. Cracks start forming.

I called this back in the spring. I said this whole thing cracks open in July. I was early, and I’ll own that. The dates were off, the diagnosis wasn’t. What I’m seeing now lines up with what Webb’s hearing straight from the C-suite: this doesn’t crash in one dramatic quarter. It erodes first, then it breaks. My new call is Q2 of next year. That’s when the missed targets pile up high enough that boards stop nodding along and start asking where the return actually went.

Here’s what makes this different from 2001, and worse. The dotcom bubble was speculative capital chasing an idea that hadn’t proven itself yet. This is proven capability sitting inside companies that have no idea how to harvest it. Webb asks CEOs a simple question: if AI freed up 10% of your capacity tomorrow, where would you put it? Nobody has an answer. That’s not a technology problem. That’s a leadership problem wearing a technology costume.

No CEO got hired because they’re an AI expert. They got hired because they’re good executives, and now they’re being asked to make high stakes calls on a technology that doesn’t behave like anything they’ve managed before. Add in the fact that the people most likely to feel behind on AI are also the most senior, most tenured, and most expensive people in the building, and you’ve got a recipe for decisions that aren’t wrong exactly. They’re just weird. Webb’s phrase for it stuck with me: this isn’t the economy taking a hit, it’s the economy making strange decisions. An economy-wide hallucination, basically.

The companies that make it through this aren’t the ones with the most pilots running. They’re the ones who stop treating AI like abundance and start treating it like a cost center that needs a real owner, real integration, and a real answer to Webb’s question. Somebody has to have the job of harvesting the gains, not just generating them.

Cracks are showing right now if you know where to look. Q2 of next year is when they turn into something you can’t unsee.

Source: Nick Lichtenberg, “Why the AI economy is like a bad dating app,” Fortune, August 30, 2026.