What does AI mean for my job?

It is becoming a difficult question to avoid. What does AI mean for my job?

For some people the worry is quite direct. They wonder whether the work they do today will still need a person doing it in five or ten years. That is an understandable concern.

There are plenty of confident predictions about what AI will do to employment. Some suggest enormous job losses. Others argue that AI will create new kinds of work and make existing jobs better. The difficulty is that nobody really knows yet. We are still early enough in this change that certainty, in either direction, should probably make us cautious.

But something smaller is already worth noticing. AI does not have to take someone's job to change it.

Think about an ordinary office job. A person might spend part of their week writing emails, preparing reports, looking through documents, summarising meetings, researching questions and putting together first drafts. Increasingly, AI can help with some of those things.

At first that can feel like a useful extra pair of hands. A report that once took an hour to draft might take twenty minutes. A long document can be summarised before somebody reads it properly. Notes from a meeting can appear almost immediately after the meeting ends.

The employee is still there. Their job title hasn't changed. Their desk hasn't disappeared. But slowly, the balance of their work begins to move. Someone who used to write the first draft now checks a draft produced by AI. Someone who used to search for information now checks the information AI found. Someone who dealt with all customer questions may increasingly deal only with the unusual ones.

The job has not gone. Parts of the job have. That distinction matters.

Not every task we do at work is particularly valuable or enjoyable. Most people can probably think of something they would happily hand over. Repetitive paperwork. Routine emails. Copying information between systems. Producing the same basic report every Friday afternoon.

If AI removes some of that work, there is an obvious benefit. People may have more time for the parts of their jobs that require judgement, conversation, experience or simply understanding another person. That is the reassuring side of the story, and it is a real possibility.

But there is another side that is easier to miss. Some of the boring work was also where we learned.

Imagine someone starting their first job as a junior analyst. They probably aren't given the hardest problems on Monday morning. They begin with smaller things. They gather information. They read documents. They check figures. They prepare summaries. They produce first drafts that somebody more experienced reviews.

Some of that work can be repetitive. It can also be how they learn. After doing it hundreds of times, they start noticing when something doesn't look right. They learn which information matters and which doesn't. They make mistakes and discover why they were mistakes. Eventually, they become the experienced person reviewing somebody else's work.

AI creates an interesting question here. If more of the beginner's work is done automatically, where does the beginner get the experience?

There may be perfectly good answers to that question. Jobs may develop different ways of training people. New tasks may replace old ones. People may learn alongside AI rather than by doing everything themselves. But we shouldn't assume that will happen automatically. The first rung of a career ladder can look fairly unimportant when you are standing near the top. It looks rather different when you are trying to get onto the ladder.

There is another possibility. AI might remove repetitive work only to replace it with a different kind of repetitive work. Checking.

If AI prepares the report, somebody may still need to make sure the report is correct. If it summarises the contract, somebody may need to check what was missed. If it drafts the customer response, somebody may need to decide whether that response should actually be sent.

That can be quicker than doing everything from scratch. But it changes the person's role. They move from doing something to supervising something. And supervising work you did not produce yourself can sometimes be surprisingly difficult. You need enough knowledge to recognise when the answer looks convincing but is wrong.

That means human experience may become more important in some places, not less. Which brings us back to the same awkward question. How do people gain that experience if AI increasingly performs the work through which they once acquired it?

It would be comforting to finish by saying that AI will not take people's jobs. We cannot honestly say that. Some jobs may disappear. Some organisations may employ fewer people to perform particular kinds of work. Other jobs may change substantially. New kinds of work will probably appear too.

And the effect will not be the same everywhere. A nurse, accountant, warehouse worker, designer, software developer and restaurant manager do very different things. Talking about all of them simply as jobs affected by AI can hide more than it explains.

What we can see more clearly is that employment does not have to disappear for AI to have an effect on working life. The smaller changes matter too. A task disappears here. A responsibility moves there. Something that used to require an hour now requires ten minutes and a careful check. Gradually, the job someone is doing becomes slightly different from the job they originally learned.

Perhaps that is where much of the change will be noticed first. Not when somebody arrives at work and discovers that AI has taken their job. But when they realise that, little by little, AI has changed what their job actually is.

For now, that may be the more useful thing to watch.