Then I walked into the room and something different happened.

Not better or worse than the plan. Just real. The kind of real you only get when four thousand people are in front of you and the argument is suddenly accountable to their faces.

Here’s what I learned.

The diagnosis isn’t the hard part.

The relationship slide - two people on a sofa, backs turned, each staring at their own screen - got the biggest laugh of the day. I didn’t say a word. I just left it there. And the room erupted.

Not because it was funny. Because it was true. Everyone in that room already knew the problem. The gap between the technology conversation and the human one isn’t abstract to IT leaders. They live in it. They are the gap.

What they were hungry for wasn’t the diagnosis. It was permission - and a practical way through.

Shadow AI is a signal, not a problem.

I asked who in the room had a shadow AI problem and as we predicted - every hand went up. Then I told them they were giving the wrong answer.

You don’t have a shadow AI problem. You have an organisation full of people who want to do better work and aren’t waiting for permission. The elephant paths are being worn into the grass in real time. That’s not a governance failure. That’s your people telling you something. The job is to make the elephant path the design - not to block it.

The room went quiet in a way that felt like recognition rather than disagreement.

Monday morning is the real test.

The ending landed differently than I expected. Not because of anything I did - because of what was waiting for them. When I talked about walking back into the office on Monday, into the board meeting, into the leadership team, every face in the room shifted slightly. Because that’s the moment the argument stops being theoretical.

That’s the only measure that matters. Not what you think in the room. What you do when you leave it.

What the corridor taught me.

The conversations after the keynote were where the argument got really interesting. A recurring theme: how do you reset, reframe and rebuild the relationship between IT and the business when the drift has been running for years in both directions? The colleague failure runs both ways - IT and the business have been equally guilty of staying in their lanes. The sofa isn’t just a metaphor. It’s a management model.

One of the sharpest developments came from a conversation about language. In the manifesto I argued for moving from “no” to “yes, but.” Someone pushed back: “yes, but” still sounds like “no” if the power dynamic hasn’t changed - the qualification is still IT’s to make, unilaterally. They were right. The better formulation is “yes, and.” Not a concession. A contribution. Yes, I understand what you’re trying to achieve. And here’s how we get there in a way that actually works. That’s not a communication style. That’s a fundamentally different understanding of what the job is.

And underneath all of it, the question about time. A CIO said to me: I buy all of it. But how do I take a board that measures everything in quarters and get them to think in years? The efficiency trap isn’t just about what we measure. It’s about when. We are trying to deliver long-term transformation with short-term political and economic tools - and that mismatch is precisely where most transformation programmes go to die.

The need has never been greater.

Here is the thing I keep coming back to. The need for the IT function has never been more urgent. In a world of autonomous agents acting on behalf of the organisation, the people who understand both what the technology makes possible and what the business actually needs are the most valuable people in the room.

But not in the way, shape or form that function currently exists.

The cost centre model isn’t just insufficient. It’s the wrong answer to the right question. The navigator role isn’t a nice upgrade. It’s the only version of this job that is fit for what’s coming.

The right question is still the right question. Vegas just made the stakes clearer.

If you haven’t read the manifesto yet, it’s here. And if you have - I’d love to know which parts landed for you, and which parts you’d push back on.