History doesn’t repeat itself, but it does rhyme. And we’ve seen this story before. As AI matures, adoption will increasingly require more than signing up for a ChatGPT account or, for businesses, buying hundreds.
That’s not transformation. That’s giving everyone a new tool. The next stage requires redesigning how work gets done. Adding AI to existing workflows can make them faster and cheaper. But there’s a ceiling to optimizing yesterday’s way of working.
The bigger opportunity is to question the workflow itself:
- Why does this process exist?
- Why are these steps necessary?
- Why does this person hand something to that person?
- Why does this approval exist?
- Why does a human need to do this at all?
And if you’re redesigning work, you can’t design for where AI is today. You have to design for where it’s headed. Where is that? Toward organizations where AI agents may vastly outnumber people.
That requires a completely different operating model.
Today, humans carry enormous amounts of organizational knowledge in their heads. They know how things actually get done, who to call, what exceptions matter, which rules can be bent, and all the little pieces of context that never made it into the process manual.
AI agents don’t automatically have that. They need context. Goals. Rules. Access. Memory. Tools. Feedback. Evaluation. Which means humans increasingly move from doing the work to orchestrating it. We become architects, managers, and judges of increasingly capable digital workers.
And let’s be clear about where we are today. Most businesses aren’t deploying armies of AI agents. They’re chatting with their favorite LLM. But that won’t be the end state.
In the not-so-distant future, we won’t spend our days prompting AI back and forth to complete individual tasks. We’ll increasingly delegate outcomes. And that’s where the history of innovation starts rhyming.
Incumbents almost always respond to technological shifts by using the new technology to improve the old model. Startups don’t have that constraint. They don’t have legacy workflows to protect. They don’t have organizational structures built around yesterday’s technology. They don’t have reputations tied to the old way of doing things.
They get to start with a blank page. That’s an enormous advantage.
Meanwhile, most established businesses are run by professional managers whose job is to grow, optimize, and protect the existing business; not destroy it and build the next one. That’s the trap.
AI makes the existing business better, so everyone celebrates the productivity gains while someone else is quietly asking whether the existing business should work that way at all. History doesn’t repeat itself. But it does rhyme.
So if you’re leading an established business, your job isn’t merely to adopt AI faster. It’s to find the revolution before the revolution finds you. Because it will.
Start with one uncomfortable question: If we were starting this company today, with AI at the core and none of our existing processes, systems, roles, or assumptions to protect, how would we build it?
Then compare that company to the one you’re running today. The gap between the two is your AI transformation roadmap.
