AI Won't Replace You. It'll Make You Dangerous.

You've heard the line by now: "AI won't replace you — but someone using AI will."

It's true. It's also too small.

It frames the biggest capability shift of our lifetime as a defensive problem — keep up or get eaten. Fear as a productivity strategy. And fear will get you to adopt tools, but it will never get you to build anything worth keeping.

Here's the frame I take on stage instead: AI is not a threat to the called. It's an amplifier for them.

The keynote version, delivered live in Nairobi at the IHA conference.

The question underneath the question

When someone asks me "will AI take my job?", they're rarely asking about employment. They're asking: does what I carry still matter?

And the honest answer is: it matters more than it ever has.

For most of history, the constraint on a person with conviction was capacity. You could see the whole problem — the kids falling through the school system, the patients lost in the intake process, the small businesses that die for lack of back-office — and you could touch almost none of it. Vision was cheap. Execution was expensive. So the people with the deepest convictions spent their lives doing a fraction of what they saw.

AI collapses that constraint. Not the conviction — the capacity.

Conviction + infrastructure

Everything I build sits on one equation:

Conviction without infrastructure is noise. You feel called, you speak, you post, you burn out — and the thing you saw never becomes a system anyone else can run. Infrastructure without conviction is soulless scale. That's most of what the AI economy is producing right now: faster funnels, cheaper content, optimization of things that never deserved to exist.

But conviction with infrastructure? That's a person who can do a generation's worth of work in a season. A workforce program that used to track thirty students on spreadsheets now coaches three hundred with the same staff. A ministry that answered questions one conversation at a time now answers them at midnight, in two languages, without losing its theology. A founder who used to need a team of twelve needs four — and the four do the work that requires a soul.

That's what "dangerous" means. Not dangerous to your neighbor — dangerous to the problems you were sent to solve.

The gap nobody talks about

The AI conversation is stuck arguing tools — which model, which app, which prompt. That's the smallest part of it.

The real divide I see, sitting with executives and pastors and founders across three continents, is this: most organizations are not lacking effort or tools. They're lacking operating infrastructure. Disconnected inboxes, spreadsheets, forms, CRMs, calendars, and manual follow-up — and an exhausted team holding it all together by heroics.

Drop AI on top of that and you get faster chaos.

Redesign the operating layer underneath — your data, your workflows, your team's actual decisions — and then AI becomes what it was always going to be for you: leverage on everything you already believed.

Build what scales — without losing your soul

There's a version of scale that costs you the very thing you were scaling for. Everyone has watched it happen: the organization gets bigger and hollower at the same time.

So the standard I hold — for myself and for everyone I build with — is double: build what scales, without losing your soul. The system carries the load; the person keeps the call. Both, or it isn't finished.

That's the work. Not adopting AI so you don't fall behind — installing it so that what you carry finally moves at the weight it deserves.

Where to take this

• If this is where your head is at, the weekly letter below goes one layer deeper — one conviction, one system you can install.
• If your organization is the one holding it together by heroics, start with the builds.
• If the question is less about your organization and more about your own next season — that's a different door.
• And if you want this argument in your room, the keynote travels.