Learning AI to leave a corporate job

I expected this work to be about productivity. Faster decks, cleaner models, an inbox that behaves. Some of it is.

But more than half the people who come to me are not trying to get better at their job. They are learning AI to leave a corporate job, or, more precisely, to reach the point where leaving is something they could choose. They rarely say it in the first session. It comes out in the third, usually sideways, usually after we have spent an hour on something that has nothing to do with their employer.

Why people end up learning AI to leave a corporate job

The trigger isn't always the work itself. It is the discovery that the work is not the part they control.

I spent thirteen and a half years in institutional banking, first in Australia and then in Hong Kong. I watched good people leave in restructures that had nothing to do with them. Teams were cut because a strategy deck in another country changed a line item. People who had given a decade to a firm found out on a Tuesday morning, in a room booked for fifteen minutes.

The question people ask out loud is whether AI is going to replace their job. What actually removes them is duller than that, and has nothing to do with technology.

Nobody is made redundant for being bad at their job. They are made redundant because a line on a spreadsheet moved.

Once you have seen it happen to someone competent and loyal, you stop believing that competence and loyalty are a plan. They are good qualities. They are not protection, and treating them as protection is how people end up with no options at forty-five.

The same machinery fails in the other direction, and that half is harder to say out loud. We have all seen it. Someone who is not good at the job, or who is good at it and dishonest with it, and who remains at the firm.

It is worse when it is the layer above you. A lot of people are not trying to get away from the work. They are trying to get away from leadership that cannot make a decision, or that makes bad ones with complete confidence, and that will still be in place next year because nothing in the structure removes it.

So the lesson arrives twice. Being good is not protection. Being bad is not disqualifying. Once you have watched both happen in the same building, you stop reading the organisation as a meritocracy that occasionally misfires, and start reading it as something that was never sorting on merit to begin with. That is usually the moment a person starts wanting something of their own, and at a minimum knowing they have the option to leave whenever they want.

So the interest in AI is not really an interest in AI. It's freedom, it's options. It is an attempt to build something that belongs to them: a skill their employer did not give them, cannot take back, and does not appear on an internal competency framework.

Having a job is not the problem

This topic attracts a lot of nonsense. There is a genre of online advice that treats employment as a failure of nerve. Quit, go independent, escape the matrix. It ignores that bills exist, that mortgages are real, and that a lot of corporate jobs are genuinely good.

A corporate job is fine. Many are better than fine. The salary is real, the medical cover is real, and the work can be interesting.

The problem is not having a job. It is being stuck in one you no longer want, with no realistic way out.

A hard week is not the same thing, and you recover from a hard week. What does not lift is the feeling that the decisions that matter are not yours to make. That is the part people carry home with them.

It's about the option, not the exit

Most people do not leave.

They may build something on the side, or they learn to run a piece of their own work without needing three other departments, and then they stay. The job becomes tolerable again, sometimes genuinely good again, because it is now a thing they are choosing rather than a thing they are stuck inside. The resentment that was building quietly for two years stops building.

That is not a consolation prize. A person who could leave behaves differently at work: they push back in meetings, they take the assignment nobody wants because it interests them, they stop managing their manager's moods. The option changes the relationship even when it is never exercised.

And when someone does leave, the difference is that they left on a date they chose, with something already running, rather than on a Tuesday morning in a fifteen-minute room.

The part nobody in banking talks about

Skills are one leg of an option. The other is money, and it is missed by the people you would least expect.

Across both offices, the number of colleagues with no investments of their own was genuinely surprising. People who could take apart a structured product for a client on a Tuesday and had never bought anything for themselves. The salary arrived, the lifestyle grew to meet it, and the balance finished each month where it started.

That is a trap, and a well-paid one. A large income with nothing behind it is not freedom. It is a subscription you have to keep paying, and it looks like success right up until the salary stops.

Getting out took longer than the short version suggests. Thousands of hours before I left: lunch breaks, evenings, weekends, commutes, a good part of most holidays. Investments, digital assets, history, economics, geopolitics, global liquidity. Reading at the wrong end of the day while everyone around me did something more enjoyable, with no evidence any of it would matter. Those hours bought the hours I have now.

I am not a financial adviser and none of this is advice about what to buy. It is the observation that financial literacy is the second half of having a choice. Someone with assets behind them has a different relationship with their employer than someone on the same salary with nothing. Not because they are about to walk. Because they could.

I have not applied for a job since I left. Some people took that badly, which I have come to read as being about their position rather than mine. I decide how many hours the work takes. The rest go on training, travelling, investing and getting better at what I do.

Where this lands

The question worth asking is not whether AI will change your industry. It will, unevenly, and the predictions are mostly noise.

The narrower one is better: if your role disappeared in the next restructure, what would you have that belongs to you?

If the honest answer is "nothing yet", that seems worth fixing. Not because your job is bad. Because otherwise you find out on somebody else's timetable, in a room booked for fifteen minutes.

I am not arguing that anyone should leave. I am arguing that you should be able to.