Technology doesn’t move forward — it moves in cycles.

The conventional wisdom about technology is wrong.

We talk about technology as if it moves in a straight line — each innovation building cleanly on the last, progress compounding into an ever-better future. But that’s not what history shows us. Technology moves in cycles. Massive, structural cycles that don’t just improve how we do things — they rebuild what economic value looks like from the ground up.

We’re at the start of one of those cycles right now. And most businesses are treating it like an upgrade.

Every 25–35 years, the input to production changes fundamentally.

Not incrementally. Fundamentally. The steam engine didn’t make hand-craft slightly more efficient — it restructured the entire relationship between labour, capital and scale. Electrification didn’t make gas lighting better — it made the economics of production unrecognisable to anyone who had built around gas.

Each wave follows the same pattern. A new general-purpose technology emerges. Early adopters experiment. Most businesses layer it on top of their existing models. Then, gradually, the businesses that rebuilt around the new input pull so far ahead that the others can no longer compete on the same terms.

The businesses that broke weren’t usually bad businesses. They were businesses that defined their value by the input that was being replaced.

Every structural technology cycle produces the same split: those who rebuild around the new input, and those who use it as a tool. The gap between those two groups compounds over decades.

We are inside the intelligence cycle. It started in 2022.

The digital revolution made information the primary input to production. Data, connectivity, software — whoever could process information faster and cheaper won. That cycle ran from roughly the mid-1980s to the early 2020s.

The intelligence cycle is different. The primary input is no longer information — it’s decision-making. Cognitive work: the work of analysing, creating, communicating, and deciding. Work that, until very recently, could only be done by trained humans at significant cost.

That cost floor is collapsing. Not for all cognitive work, and not all at once — but for a widening category of tasks that currently represent the majority of what expert-led businesses sell.

This cycle will run for approximately 25–30 years. The businesses that rebuild around it will define the next economy. The businesses that use it as a tool will be outcompeted by the ones that did.

The businesses at risk are the ones that sell cognitive labour at a premium.

If your business model is essentially “you pay for our people to think, create and decide for you,” you need to look very carefully at what you’re actually selling.

Not because AI will replace your people. But because the cost structure of delivering that cognitive work is about to change for your competitors. And when your competitor’s cost floor drops significantly below yours, one of three things happens: they undercut your pricing, they reinvest the margin into quality you can’t match, or they scale beyond where you can follow.

In every previous cycle, the businesses that broke were the ones that held on longest to the old input. Not because they were slow to adopt the technology — but because they never rebuilt their model around it.

Systems builders. Always.

The pattern holds across every cycle: what wins is the businesses that treat the new technology as infrastructure, not a tool. Not “we use AI” — but “our business is an intelligence system.”

This means building processes that compound. It means documenting what you know and converting it into systems that can run without you being present for every decision. It means treating your expertise as an asset to be systematised, not a service to be delivered manually hour by hour.

The digital winners weren’t the ones who had computers. They were the ones who rebuilt their businesses as digital systems. The intelligence-cycle winners won’t be the ones who use AI. They’ll be the ones who rebuild their businesses as intelligence systems.

You have a window. It’s not infinite, but it’s real.

Right now, the cost of restructuring around AI is relatively low. The tools are accessible, the competitive pressure is building but not yet decisive, and most businesses in your category haven’t made the move yet.

In 28 years, asking whether your business uses AI will be like asking whether it uses the internet.

The founders who understand cycles — who can see far enough ahead to build toward a future that isn’t consensus yet — are the ones who win the transition. Not because they were smarter. Because they moved while the window was still open.

That’s what 2050 Expert is built for. If you want to understand what the intelligence cycle means for your specific business — where you’re exposed, where the opportunity is — start with the Commerce Audit. It’s the fastest way to see clearly.