Society has extreme inertia

Society has extreme inertia.

I’ve had 896 conversations with ChatGPT since September 2023. I build projects with Claude Code. I’ve created custom agents, automated workflows, and analyzed my own GDPR data exports with AI tools most people don’t know exist.

If there’s a spectrum of AI users, I’m at the far edge.

And I think most companies will barely notice AI five years from now.

That’s not a contradiction. It’s an observation about how organizations actually work.

Start with banks. Nobody switches banks. Not because the bank is good, but because switching is a hassle. Swedbank can have scandal after scandal. Customers stay put.

It doesn’t matter if an AI assistant can answer questions ten times faster. The threshold for change in a banking relationship is extremely high.

Take Volvo. Ericsson. Atlas Copco. Large industrial companies with tens of thousands of employees. Do they even want employees to become dramatically more productive with AI?

No. The organization isn’t built to absorb individual productivity gains. An engineer who suddenly delivers three times faster creates no effect if the approval process still takes six weeks. The system’s inertia absorbs the individual’s speed.

There are exceptions. Small companies, IT firms, consulting firms with short decision paths. AI can make a difference there already.

But they represent a small part of the economy. Most people don’t work at tech companies. They work at municipalities, regional governments, industrial companies, logistics firms, construction. Organizations with long implementation cycles, heavy regulation, and change fatigue.

Within five years, I think ordinary people might start using Projects in ChatGPT. That’s roughly the level.

Not because the technology can’t do more. But because adoption requires behavior change, and behavior change requires incentives. Most employees have no incentive to change how they work. They have incentive to do what they’ve always done, without causing problems.

The fault lies neither with the technology nor with the individual. The organization works the way it was designed to work.

Every time a new technology is introduced, the same pattern appears. The technology can do X. The organization absorbs 10 percent of X. The rest gets stuck in processes, decision paths, skills gaps, and lack of ownership.

It was true of ERP systems in the 90s. It was true of cloud services in the 2010s. It’s true of AI now.

There’s a test. Ask ten employees at a mid-sized industrial company if they use AI in their work. Not if they’ve heard of it. Not if they’ve taken a course. If they actually use it, daily, in their job. The answer at most places is zero out of ten. Or one out of ten who experiments privately but doesn’t tell the boss.

That’s not because people are hostile to technology. It’s because nobody asked them to change. Nobody measures it. Nobody rewards it. Nobody has explained what it’s supposed to replace.

And in an organization where 90 percent of work is governed by processes written in 2018, it’s not strange that a new tool doesn’t find its place.

I’m not saying AI doesn’t change things. It does. I’m saying the change happens slower, more unevenly, and more unpredictably than the people selling AI services want you to believe.

The most honest thing I can say as a power user: the technology works. Organizations’ ability to use it does not. That combination doesn’t produce exponential change. It produces slow, uneven, context-dependent adoption.

That’s less catchy as a headline. But it’s closer to reality.