There is no switch

I have been following this for 3 years myself. I still rewound the video a couple of times.

The timeline was not the reason. The reason was a detail he was careful about: it will not arrive as an announcement.

My take is that models improve month by month, and each release does something the last one could not. Alongside that, AI-powered robotics is moving fast and quantum computing keeps developing. Put those three together and you are describing a trajectory rather than an event.

Which means the thing people are waiting for has already started.

Nearly every senior professional I speak to treats this as binary. Either AI takes my job or it does not. Either the predictions are right or they are overblown.

That framing gives you permission to wait. If it is binary, there is a moment when you will know, and until that moment you can carry on as you were.

There is no moment. There is a slope, and we are already on it.

Here is where it starts, and it looks like nothing. In the past few years I have been asked more than once, in senior commercial meetings, how we might use AI to avoid recruiting new headcount and get more from the people we already have. Nobody in those rooms described it as job losses. It was framed as productivity. It was framed as good management.

That is what this looks like from the inside. No redundancy announcement. A budget conversation where a role that would have traditionally been created simply never gets opened.

Roughly half of layoffs now reach manager level and above (in the US and UK). Manager headcount at large corporates fell 6.1% over three years. Those numbers are the visible end of conversations that began much earlier.

I could be wrong about the timeline. So could the researchers. Nobody knows.

But two questions matter regardless of which future turns up.

What do I want to do next.
And how do I get there.

If the predictions are right, those questions matter more rather than less. Work may become optional for a lot of people. Identity will not. A great deal of what people do is who they are, and very few will choose to stop entirely. In that world, what you actually want to spend your time on, and whether you can make it pay, becomes the whole question.

If the predictions are wrong, you have spent two years upskilling, building optionality, and reducing your dependence on a single source of income. Nobody regrets that.

There is no version of this where preparing is wasted. Which is why I am comfortable making the argument even with an uncertain timeline.

This next part is my reasoning rather than anyone's forecast.

If you are late in your career and treating your pension as the settled part of the plan, examine that again. Pension promises assume a tax base and an economy that look broadly like today's. If the disruption is anywhere near what is being described, that assumption deserves a second look.

I am not saying it fails. I am saying it is an assumption we never had to test so far.

Open whichever AI assistant you already pay for and create a persistent project inside it. Not a chat you abandon. A project, with context you build up over time.

Then give it something real. A commercial problem you are working through. A decision you are stuck on. Ask it to challenge your thinking rather than agree with you.

Two reasons for this. First, you learn what the technology can actually do, which is different from what you have read about what it can do. You find out where it is strong, where it invents things, and what quality of input it takes to get output you would put your name to.

Second, you build a working understanding of the thing reshaping your industry. Knowing it from the inside is a different order of knowledge from reading about it.

One caution. Use it to think with, never to think for you. Outsource your judgement and you lose the ability to make the calls that make you valuable, and that erosion is slow enough that you will not notice it happening.

My thinking on this has moved a long way in the past year, and it kept pulling at the title.

This letter used to be called Leading in the Age of AI. That was a leadership newsletter, and leadership is no longer doing the subject justice. What I want to write about is what this technology is doing to work itself, and what senior professionals can do about it while there is still room to act.

So I have changed the name to The AI Disruption Brief. Same writer, same reader, wider frame, and a clearer promise: I will keep you informed on how AI is reshaping the workforce, and I will not soften it.

Take it as a working assumption that I am right. You do not have to believe it. The cost of preparing and being wrong is small. The cost of waiting and being wrong is not.

Then ask what you would do differently.

Reply and tell me what comes up.

I am building Zelova, a career strategy service for senior professionals working through exactly these questions. Early access is open at zelova.ai.