Will YOUR job be made redundant? Experts reveal the career paths that are in serious trouble – and the ones you should join as safe havens from the AI apocalypse

Today, if your career choice means you work in front of a computer screen, AI is after your job; Mustafa Suleyman, Microsoft’s head of AI, believes most white-collar work will be ‘fully automated’ within 12 to 18 months.

And a few boffins in AI are starting to frighten themselves. Some of those who walked out of ChatGPT early on because of safety concerns started a rival company, Anthropic, makers of chatbot Claude. 

This month one Anthropic developer Mrinank Sharma, who has a computer engineering degree from Cambridge and a tech PhD from Oxford, has left, saying that ‘the world is in peril’. Sharma is now planning to take a degree in poetry.

Some still say the panic is overblown – I did too until a few months ago. But it’s now very clear that AI is the most significant workplace crisis since the age of computerisation began in the 1950s. 

The Institute for Public Policy Research calculates that eight million UK jobs are under threat, from architects and accountants to those working in customer service. If you want to help any youngsters in your life make sound, future-proof career decisions, read on.

IN PERIL 

Lawyers

Shortly before Christmas, a senior London barrister gave an anonymous interview to The Spectator magazine that made headlines. 

He said he had recently spent a day working on a legal opinion – likely cost to his client of several thousand pounds – then asked AI to do the same task, which it did, in seconds, and at no cost. 

AI’s efforts were, the lawyer admitted, frankly better than his own. He was inclined to think that, thanks to AI, law was doomed as a career.

The problem, the barrister said, was not so much that a top KC could be outwitted by AI, but that entry-level lawyers would no longer be needed. And with no young lawyers to develop into senior professionals, experienced lawyers will eventually die out.

That anonymous barrister’s conclusion resonated with me. Last year, a 22-year-old friend decided to take legal action against a cousin over a family matter. 

The errant cousin had engaged one of the most expensive solicitors in London to fight off the action. My friend – a young woman – in turn had asked a medium-ranking solicitor in the suburbs to represent her. 

‘You have an excellent case,’ the lawyer said. ‘But we’d need £60,000 upfront to represent you and the final cost could go into six figures.’

Horrified, she got ChatGPT, Claude and Google’s Gemini to work on the case for her and appeared in front of the court as a ‘litigant in person’ – representing herself.

I went along to the High Court case, worried that it would be an embarrassment, but was amazed when the judge, the court clerk and a barrister sitting in on the hearing said how impressed they were with the litigant-in-person’s work – and how slapdash the expensive opposition solicitor had been.

The proceedings continue, but what I witnessed has extraordinary implications. If AI can handle such a complex case competently, something as simple as a house purchase or an uncomplicated divorce will, before long, surely be done automatically by AI in seconds.

Taxi drivers

Oddly enough, it was cab-driving that became an early victim of pre-AI technology and suggests the pattern of things to come. 

Ten years ago, the choice in London was between black cab drivers, who spent years learning routes doing the Knowledge, and untrained minicab drivers, armed with only an A to Z, who often had no idea where they were going.

Then came sat nav, then smartphone apps like Uber and, suddenly, it was cheaper, easier, quicker and more comfortable for the public to use the modern alternative to the black cab. 

The last laugh, of course, will be on both types of taxi when AI-run self-driving vehicles will make almost all human driving jobs as obsolete as blacksmithing became when the car itself was invented.

LOOKING SHAKY

Doctors

‘I wouldn’t tell a young person not to try medicine,’ a medical consultant friend tells me. ‘But it’s going to be a very different profession from now on. We forget that usable AI is only three years old. Imagine how it will be in ten years.’

He cites a key example of how things have already changed: doctors now routinely use AI to view X-rays. ‘AI will look at every pixel of an X-ray in an instant and is pretty well incapable of making a mistake,’ he says. 

‘But if you’re a human viewing a hundred X-rays, it’s impossible NOT to make the odd mistake. I’m just glad I’m retiring soon.’

The NHS is reportedly planning to let GPs work remotely full time, even if they are on a beach. But people are already consulting AI first, even showing it photos of rashes, bumps etc before seeking a hard-to-get doctor’s appointment. 

Skilled surgeons will still be needed, but, like lawyers, gaining experience will be an issue. Oh, and forget psychotherapy or psychiatry. Millions already consult AI on emotional matters. It’s free and never gets bored with your problems. You can have 12-hour ‘appointments’ if you like.

HOPEFULLY OK

Nurses, teachers, hairdressers, physios, farmers, chefs

These jobs rely on manual dexterity, and while we always thought robots would take over such work, there’s never been one that could so much as carry a cup of tea without spilling it. There are rudimentary kitchen robots, but they are more amusing than useful.

Teachers are an odd case, because even 50 years ago, future gazers like TV’s Tomorrow’s World were saying ‘teaching machines’ would soon make them obsolete. However, they were trialled but didn’t catch on. The issue? 

The machines worked for basics like arithmetic and spelling drills, but failed – as will AI – to stimulate educational essentials such as curiosity, creativity, discussion and social learning. Children quickly became bored, confused or stuck – the human touch of a good teacher can’t be replicated.

(Good) writing

While bottom-end journalism is already taking a hammering, with AI churning out fake or repackaged old news by the ton, serious investigative journalism and intelligent, lighter-touch newspaper writing is pretty safe. 

AI can summarise facts, but it can’t be curious, funny or morally courageous. High-end writers, composers, designers and filmmakers who have taste, cultural authority and a personal take on things could thrive in a market flooded with literary slop.

Entrepreneurship

Another form of creativity that should be in the ascendant is entrepreneurship. It is highly unlikely that in the near or medium future, your computer will be like the talking toaster in the old Red Dwarf science-fiction series and pop up with ideas it’s not been asked for.

Vets

I’d put a career as a vet on the list of things I’d tell a teenager to consider. Why? Because while you and I can tell an AI doctor what’s the matter, a cat can’t.

THESE SEEM FINE 

Tradespeople

Electricians, plumbers and carpenters have an assured, lucrative future. Finding and repairing a specific pipe in a specific house is impossible to automate and always will be.

Hospitality

This sector will always need people. As Martin Ford, the futurist and award-winning author of four books on AI’s threat to traditional work, says: ‘To automate jobs like these would require science-fiction robots, like C-3PO in the Star Wars movies.’

One last thing: how about going into AI itself?

Here’s my tip for a safe job in the future: if you’re made of sterner stuff than the Anthropic-developer-turned-poet Sharma, what about going into AI? 

Back when computerisation was going to take over everything and we’d be swanning around doing whatever, nobody envisioned that the IT industry would instead become one of the biggest employment opportunities of the future.

So today, when AI can do more grunt work than we ever expected, it is more than likely that the AI industry will need more and more humans to build the AI to do said grunt work.