The End of the Traditional Career?

Think about how people traditionally become experienced professionals. A junior accountant prepares financial information, a trainee solicitor reviews documents and a graduate analyst researches markets and builds presentations. The work can be repetitive and sometimes frustrating, but it gives people exposure to real problems, colleagues, clients and decisions. That exposure is how judgement develops. As I argued in the second article in this series, much of that learning comes from the effort of doing the work yourself, which is exactly the effort AI can remove.

Now imagine AI taking over a large share of those activities. A senior colleague can use it to produce a first draft, summarise documents or analyse information without passing the task to someone more junior. That may make the senior colleague more productive, but it creates an obvious problem. If AI reduces the need for entry-level work, where will the next generation of senior people come from?

There are early signs that this is more than a hypothetical concern. Researchers at Stanford's Digital Economy Lab have been tracking US payroll data from ADP, covering millions of workers. Their latest update, published in August 2026, found that employment among 22- to 25-year-olds in the occupations most exposed to AI was about 19% below where it would have been had it kept pace with less-exposed occupations.^[1] Older, more experienced workers in the same occupations showed no comparable gap. That's US data and it can't prove AI is the only cause, but it's a pattern worth taking seriously.

The picture isn't one-directional, though. The customer-support study I discussed in the first article in this series found that AI helped the least experienced workers most, improving their productivity by around a third.^[2] So both outcomes are possible. One employer might recruit fewer graduates, while another hires the same number and uses AI to help them contribute sooner. The danger lies in organisations focusing on immediate efficiency without thinking about their future workforce. A business that cuts junior recruitment today may find in five years that it lacks people with the experience needed for senior roles. You can't promote people you never hired, and knowing how to use AI isn't the same as understanding the work it produces. The ladder may survive, but employers will need to rethink how people climb it.

Imagine two professionals. One has spent ten years building expertise in a particular field. The other has two years' experience but is highly capable with AI tools. Both can now produce a polished report in minutes, but that doesn't make them equally valuable. The experienced professional is more likely to notice when the analysis is misleading, when a recommendation won't work in practice or when an apparently straightforward decision carries hidden risks. They understand the client, the organisation and the consequences of getting things wrong, and none of that comes automatically with an AI-generated answer.

Experience isn't a guarantee of future value, though. If much of your expertise lies in completing tasks that AI can now perform reliably, the market may value that particular capability less. That doesn't erase the years you spent developing it, but it does change what employers are willing to pay for.

For some experienced professionals that will be an advantage. For others it may mean an uncomfortable reassessment of where their value really lies. And for people early in their careers, it could mean finding ways to show good judgement before they've accumulated the traditional number of years on their CV.

For most of modern working life, progression has meant moving up within an organisation, from junior to senior to manager to director. Increasingly, the next move might be working for yourself instead. AI is lowering some of the practical barriers to independent work. A consultant can use it to help research a market, draft proposals, organise information and build a website. A designer can explore ideas faster, and a small business owner can automate parts of their customer administration. Work that once needed several specialist services can now be handled with a combination of AI tools and the person's own expertise.

That's a genuine opportunity, particularly for experienced professionals with specialist knowledge, who might build a small business around the expertise they already have rather than waiting for the next promotion. But AI doesn't turn anyone into a successful consultant overnight. You still need something worth selling, customers who trust you and the ability to deliver work people will pay for.

The appeal of choosing your clients, setting your own hours and letting AI take care of the admin also comes with a less glamorous side. There's no guaranteed monthly salary, no employer-funded holiday pay, no employer pension contributions and no certainty about where the next contract will come from. AI can help you prepare a proposal, but it can't make a client accept it. It can help organise your finances, but it can't make an overdue invoice arrive. And while it can support your marketing, it can't replace the trust that makes someone choose you over a competitor.

For someone with established expertise, financial reserves and a strong professional network, independent work may be a realistic and rewarding option. For someone just starting out, supporting a family or dealing with unpredictable expenses, the risks are much harder to absorb. Conversations about AI-enabled entrepreneurship tend to focus on how easy it has become to start something, and spend much less time on how hard it can be to sustain it.

The idea of a single profession for an entire working life already feels less certain than it once did. People change employers, industries and sometimes careers altogether, and AI could accelerate that movement. A marketing professional might become an independent consultant, an HR specialist might move into AI implementation, and a software developer might build a small product business. Some people will combine employment, freelance projects, teaching and a business of their own. These portfolio careers aren't new, but AI widens the range of work one person can take on with limited resources.

We should be careful not to romanticise constant reinvention, though. Changing careers can mean starting again financially, losing seniority, rebuilding professional relationships and competing with people who already have years of experience in the new field. Not everyone has the savings, confidence or circumstances to take those risks, and there's a big difference between having the freedom to reinvent yourself and feeling you have no choice.

Much of the commentary on AI and work also carries an unspoken assumption that everyone should respond by becoming more entrepreneurial, more adaptable and more comfortable with uncertainty. Plenty of people simply want a stable job they enjoy, with colleagues they value and no interest in running a business. There's nothing outdated about that. Predictable income, paid leave, pensions and a clear boundary between work and home are worth a great deal, and AI shouldn't turn stability into something people feel embarrassed to want.

Employers have a responsibility here too. When AI changes the work people do, organisations can support them by redesigning roles around what AI does poorly, creating development opportunities and helping people move into new responsibilities, rather than treating constant reinvention as an individual problem. Not every job can be preserved indefinitely, but people shouldn't be expected to navigate every change AI brings on their own.

If AI lets smaller teams achieve more, businesses might gradually rely less on permanent employment, keeping a smaller core workforce and bringing in independent specialists for particular projects. That's plausible in some areas, and AI could help those specialists deliver more efficiently.

There's a limit to how far this logic goes, however. Companies don't employ people only because tasks need completing. They need continuity and people who understand their customers, systems, relationships and history. They need employees who can respond when circumstances change, take responsibility for decisions and work on problems that don't fit neatly into a project brief. Much of the value a permanent employee provides doesn't appear on a list of completed tasks. Relying heavily on contractors brings its own problems too, because knowledge walks out of the door when assignments end, coordination becomes harder and accountability becomes fragmented.

So I don't think AI points towards a future without permanent employees. A more likely outcome is that organisations experiment with different combinations of permanent staff, contractors, specialist partners and AI-supported processes, with the balance varying by industry. That alone could change what career progression looks like, even for people who stay employed throughout their working lives.

Ask someone what they do and they'll usually tell you their job. "I'm a teacher", "I'm an engineer", "I'm a nurse". Those answers carry far more than information about income. They represent years of effort, professional relationships, personal achievement and a sense of belonging. If AI changes the value of particular skills or makes familiar career paths less predictable, the consequences won't be purely financial, and some people will need to rethink how they see themselves.

That could be liberating. Someone stuck in a profession they no longer enjoy might find new opportunities, and someone with specialist knowledge might finally have the tools to build a business of their own. It could also be deeply unsettling, particularly for people who have spent much of their working lives becoming good at something. We shouldn't underestimate what it means to be told that the profession you've spent twenty years mastering now needs to change, or assume that everyone will welcome that change simply because technology makes it possible.