You need experience to get a job. But you need a job to get experience.
It’s a familiar frustration. Now add another complication: what happens when AI starts doing the tasks you would have learned on?
There are concerns about AI’s possible effect on graduate opportunities.
For businesses, that raises a question worth thinking about: could saving money on junior staff today leave them short of experienced managers tomorrow?
Those “boring” tasks have a purpose
Ask an experienced accountant about their first job and you probably won’t hear about negotiating a takeover on day one.
More likely, they checked invoices, reconciled accounts and updated spreadsheets.
Hardly the opening scene of a Hollywood blockbuster.
But those tasks teach you things. You learn how transactions flow through a business, why records don’t always match and what an unusual entry looks like.
That experience helps develop professional judgement: knowing when something makes sense and when to ask another question.
If software takes over more of that work, businesses need to think about how beginners will develop the same understanding.
The attraction is obvious
Imagine a finance team using AI to extract information from invoices and flag potential errors.
If it works reliably, staff can spend less time processing information and more time investigating problems. That could mean faster work, lower costs and fewer repetitive tasks.
For a CFO or finance director under pressure to improve efficiency, that’s an attractive proposition.
However, automating a task and removing a role are different decisions. A junior employee might spend less time entering data and more time checking exceptions, discussing discrepancies with suppliers or helping explain results.
The job changes. The learning needs to change with it.
A saving today, a shortage tomorrow?
Here’s a hypothetical boardroom decision.
A company reduces its graduate intake because technology can handle more routine work. Its salary bill falls, and the immediate financial results look better.
Several years later, it needs experienced team leaders. But fewer people have come through its own training programme.
It can recruit externally, of course. Yet if other employers have made similar cuts, experienced candidates may become harder and more expensive to find.
This is a classic short-term versus long-term planning problem. The saving is easy to measure now. The cost of a weaker pipeline of future managers may take years to appear.
That doesn’t mean every reduction in recruitment is a mistake. It means recruitment decisions need to look beyond this year’s budget.
So, how do people learn now?
Businesses could redesign junior roles to combine technology with deliberate training.
A trainee might review an AI-generated analysis, explain its conclusions to a manager and investigate anything that looks odd. Rotations between teams could help them understand the wider business, while supervised exercises could build skills they no longer practise through routine processing.
There’s a catch, though: you can’t meaningfully check an answer if you don’t understand the subject.
Technical knowledge still matters. A polished explanation or a neatly formatted spreadsheet can contain mistakes.
For students, that means learning the fundamentals, becoming comfortable with technology and practising how to explain your reasoning. Being able to ask “Why has this happened?” is useful. Being able to judge whether the answer holds up is even better.
Who trains the next finance director?
AI could make junior work more interesting by reducing repetitive tasks. But that benefit won’t automatically create a good training programme.
Employers still need to give beginners opportunities to practise, make supervised mistakes and develop confidence.
After all, the next CFO is probably starting their career somewhere today.
If businesses want experienced people tomorrow, how will they help them gain experience today?
