Aug 22, 2025
Managers just can't stop talking about AI. Every day, in every investor deck, at every conference. Banks claim to have hundreds of use cases. Consumer brands call AI their "new operating system." Venture capitalists just keep pouring billions into data centers as if they were building the pyramids.
But what actually happens then? Well, maybe 10% of companies are using AI in any meaningful way. Ten percent. That’s not exactly a revolution; it’s hardly a pilot project.
This shouldn't surprise anyone who knows a bit about technology history. Do you remember when Robert Solow said in 1987 that you could see computer usage everywhere except in productivity statistics? He hit the nail on the head. Every breakthrough goes through this. The dot-com bubble had the exact same tune. All that hype, all those promises, and then it took years before anyone figured out how to actually make money from it. The real gains only appeared after everyone had rewritten their routines, retrained their staff, and basically rebuilt the entire operation.
The thing is that companies are hopelessly bad at change. Sure, the CEO might be rambling on about AI during investor calls. But who decides what actually happens? The middle managers who have been running the same routine for ten years. The lawyers who raise liability questions that no one understands. HR who panics about what happens to jobs. Compliance who lies awake fretting about privacy regulations that don’t even exist yet.
Organizational inertia, you know. And it’s winning right now.
But wait, it actually gets more interesting. Some of the gains are already happening; we just don't see them. The Washington Post called it "dark leisure," and it’s actually brilliant. The lawyer who uses ChatGPT to draft contracts in 20 minutes instead of three hours? They don’t tell the boss anything. They just take time off on Friday afternoon instead. The efficiency is real but invisible until the managers realize that they need to completely rethink what constitutes a normal workload.
They will do that, sooner or later. They always do.
Do you know what’s really ironic? The adoption feels slow precisely because the potential is so damn huge. Technologies that really change industries never come in neat quarterly reports. They spread oddly. Some departments jump on them immediately, while others battle them for years. Do you remember spreadsheets? People mocked them as nerd toys. Then came VisiCalc, Lotus 1-2-3, and Excel and changed not only accounting – they changed how every company on Earth thinks about data.
AI is on the same path. I’d bet money on it.
💡 The lesson for B2B tech companies: This delay is not failure; it’s opportunity. While everyone else is waiting for the "real" AI revolution, you can run experiments, clean up your data, find use cases that actually work. Because when the momentum finally arrives (and the productivity statistics catch up), you want to be the company that has been using this for years, not the one that is racing to keep up.
The slow march is not stopping. It has barely begun.