In college I played football with a guy who strutted, talked, and promised greatness every Saturday. One small problem: he couldn't play. One day my position coach watched him, smiled, and said, "You've gotta love confidence without evidence."
I think about that line a lot in today’s AI markets. The predictions are enormous. The evidence is not. For example:
Corporate performance. Outside of tech, the impact of AI isn't showing up in P&Ls or industry rankings.
Jobs. Three years in, there's no discernible displacement. An economist back from a desert island would scan the employment numbers and shrug.
Tech Services. The CEOs talk revolution, but the income statements don't. Infosys just touted rising AI revenue while cutting growth guidance to 1%.
ROI. Hundreds of billions are pouring into frontier labs, hyperscalers and data centers, with talk of a SpaceX-Tesla "AI supercompany." The return remains invisible.
Catastrophizing. AI doom is now a growth industry of its own. Yet it's hard to fear a technology that still can't track your overnight package.
All of that said, I'm still a strong believer in the coming AI transformation. Call it confidence without evidence. This is a slow-slow-all-of-a-sudden market.
What triggers "all of a sudden"? History's answer is a recession. Economists Nir Jaimovich and Henry Siu found that nearly 90% of U.S. job losses in routine occupations since the mid-1980s came within a year of a recession. Companies don't automate when they can. They automate when they must. In good times, layoffs are a reputational risk. In bad times, they're called "discipline" and the board applauds. Same thing is true with corporate strategies and customer preferences; times of stress force the big changes.
That's why this moment is so interesting. Much of the hard work is done: the models work, many workflows are now mapped, and the infrastructure is going up. What's missing is a forcing function. When sales drop and the stock takes a hit, the CFO will pull the AI business case out of the drawer, and pilots will become mandates quickly.
My old coach would appreciate the irony. AI's confidence was simply early. Its evidence is waiting on a bad quarter.
The week's most important AI news, curated for you:
