The A.I. job apocalypse is here; here's what should happen

The problem isn't that AI will take our jobs, it is that companies won't help us adapt.

Welcome to #MajorRelate with Dr Eliza Filby, helping you understand the social and economic trends that govern our lives…. If you have any suggestions for any topics you’d like me to cover in future newsletters, do tell in the comments!

Apologies for no newsletter last week, it’s the busiest time of year for the business - speaking and digital projects abound, and (not unrelated to this), I spent most of the end of last week in bed ill!

The job apocalypse is here; here's what should happen…

The problem isn't that AI will take over; it’s that companies haven’t invested in their people

We’re hearing it more and more. Graduates in search of white collar entry level jobs are in trouble. The unemployment rate for recent grads in the US is at its highest level since 2021. All the ChatGPT they’ve potentially used to get them ahead in the education market is proving an opposing force when it comes to the job market. Coders, once demanding high wages and bespoke terms, are now struggling to get through the door. Paid internships? Why should employers bother with the hassle when the work can be done at speed, 24/7 through AI? The lower rungs of the ladder, once essential for building a career, aren’t just disappearing. They’re being snapped off.

A New Age of Realism

Unsurprisingly, this is fuelling a new age of realism among young people. What began as a post-pandemic financial reckoning in 2022 has accelerated into a full-blown generational shift, catalysed by fears around AI. The idealism once associated with Gen Z and work may be evaporating. As one apprentice bluntly told me: “I don’t care about gender-neutral toilets. I just want job security and a decent wage.”

And right now, that kind of job security is being offered in one industry: defence.

But it’s not just graduates feeling the impact - we’re ALL de-skilling fast, as AI outpaces us on a timeline measured in months, not years. We can’t scroll anywhere without hearing about some monumental AI update or upgrade.

Anecdotally, it’s small businesses, not the corporate giants, that seem to be seeing the most immediate gains from AI integration. They’re more nimble when it comes to tech adoption and hiring. And let’s not forget: small businesses employ 45% of the private sector workforce in the US and nearly 60% in the UK.

If deindustrialisation was the collapse of blue-collar work - due to globalisation - then maybe today we’re witnessing de-officisation: the unraveling of white-collar life through AI- where the office is no longer central, careers are no longer linear, and professional status no longer guarantees security, meaning (or even a desk) and employees need to train for multiple careers.

How We Got Here

As with most things AI, I tend to think its greatest value lies not in its own brilliance but in how it forces us to confront some uncomfortable truths about humanity’s own shortcomings.

So let’s step back a moment and reflect on what’s happened to the white collar job market over the last forty years, because I believe it reveals where we need to go next.

Broadly speaking, from the 1960s, as we moved into an increasingly knowledge-based economy, we were told that white collar work was the future, we were told a degree was a necessity and that more of us needed to get one. The birth of a true meritocracy, they said. By the early 2000s, we were told that this now ESSENTIAL primer for professional work would become increasingly expensive and generate a debt that would stay with most of us throughout our working life. Just to be clear, there was no conspiracy here, just a clear economic shift that came with a cost. And yet, as

Anne Helen Petersen

writes in her book Can’t Even, that cost fell not on the state or employer, but really for the first time, on the individual worker who was essentially ‘served a bill’ for their own training.

Historically, in the age-old apprenticeship model of work, there was never an expectation of prior knowledge. Instead, employees (taken more on attitude than aptitude, and often personal connections) were paid to learn on the job, the business took on the responsibility for that teaching, and there was a job guarantee at the end.

As Daniel Markovitz sets out in his book The Meritocracy Trap, managers and executives never used to have degrees; they were nurtured in-house. At IBM, for example, if someone had 40 years of service, at least 10% of that time would have been spent in training. At Kodak, another American corporate stalwart, the initial training was 18 months long. Markovitz notes that the IBM training programme was so fundamental to the culture of the firm that when they got rid of it in the 1990s, ‘the shock at its headquarters were so great that the company officials asked local gun shop owners to close their stores.’

But that compact began to collapse in the 1980s, alongside final salary pensions, loyalty and the broader social contract of work. Training, once core to the white collar company, was basically outsourced to universities. By the 1990s, that investment was gone. In the US today, the average company spends less than 2% of its payroll on workforce training.

For obvious reasons, businesses in the 1990s embraced the idea of a work-ready graduate; it promised savings on training and a faster route to productivity. And yet, as Daniel Markovits points out, the social consequences were predictable: ‘shifting training for elite jobs out of the workplace and into the university changes the socioeconomic composition of the people who receive the training’ (and, yes, inevitably, who gets the jobs).

When graduates did enter the workforce, they were often met with limited learning and development programmes. These were well-meaning, perhaps, but designed to polish rather than build from scratch. High-quality training still exists in fields like engineering, accountancy, law, and medicine, but for most other sectors, development is front-loaded, then fades. By mid-career, especially for those not on the leadership track, it’s virtually non-existent.

Digitalisation and hybrid working have only made this worse: less time, less human interaction, more screen-based training often outsourced, and, of course, far less wisdom-sharing. Few workplaces truly value learning, reward those who teach, or make knowledge flow down, or let alone up the food chain, from junior to senior.

The AI Reality Check

AI is now exposing the consequences of this model.

The great degree consensus that defined my early adult life is slowly unravelling. Universities are in financial trouble. Students are questioning the point (and price) of their education.

Meanwhile, AI is quickly taking up the grunt work that grads used to cut their teeth on. If this continues at pace, we could potentially see a lot of young people over-educated, in debt, underpaid and ultimately unemployed. That is a political powder keg no politician (let alone young person or parent) should want to see.

But the past may point to the solution.

Many parents, students, and even educators realise that education needs to change; to return to being less academic, more vocational, less exam-centric, more holistic. Yes, we need more degree apprenticeships - the learn while you earn type - many of which are now harder to get into than Oxford or Cambridge. We also need to accept that the academic route suits some, not others, and that it is not the meritocratic path we assumed it was, but increasingly a path only made possible by rich parents.

And yet many have been saying this for years; the problem with education (as anyone who works in it will tell you) is that innovation is slooooow.

Companies themselves need to be responsible for managing this seismic shift. In the age when workers will need to outpace AI, businesses need to return to the apprenticeship model. But…. a learning model that never stops, ever. A business learning culture that rewards teachers as well as leaders, learning as much as productivity. That means reviving the apprenticeship model, but reimagined as lifelong.

A workplace where learning never stops.
Where teaching is rewarded as much as output.
Where your best apprentice might be 60, and your best leader 25.

If AI is going to transform our jobs every five years, we need a workforce that can transform alongside it.

When deindustrialisation happened in the 1980s/90s, entire industries, towns and communities were destroyed, not simply because that particular industry left, but because nothing was put in its place. The lesson from deindustrialisation is that the downside to unions is that they are likely to stifle innovation in favour of labour; the downside to employers is that they are likely to supercharge innovation to the detriment of labour.

Deofficisation won’t come with shuttered factories or picket lines. It will be quieter, more individual, but no less brutal.

If we expect people to retrain every few years to stay relevant, we cannot expect them to also carry the cost. Especially not when many are already in deep debt from their first round of education.

The future isn’t about more people. It’s about properly investing in the ones we keep.

A Story that stuck with me

For my book Inheritocracy (now available in paperback here) I interviewed Cole, 32, and from Hartlepool, in the North East of the UK, and listening to him gave me an insight into what that employer-employee relationship should look like in an age of AI.

Cole was fortunate to get a four-year apprenticeship straight from school. But what really makes Cole’s story unique was not only the fact that he was on a good apprenticeship scheme and had shunned the academic route, but was the way the company invested in him and his skills. The company paid for him to do a degree (so he graduated debt-free), and after nine years of working at the company, he was earning £44k. He eventually left, not for any other reason than there were no more promotional routes for him to go. In other words, he was now over-skilled for any future jobs in that particular company.

Cole is a rare breed, a meritocratic millennial where work raised him out of his working-class roots, firmly into the middle class. In his new role, Cole has doubled his salary to £100,000 a year: ‘We have a nice four-bed detached house which we can live in comfortably ... We have plenty of holidays, two cars and the kids are very fortunate...’

His employer didn’t just hire a worker. They built one.

And for once, work really did change someone’s life.

‘They were a great company; I don’t think I will work for a company as good as them ever again.’

Oh, and another reason why Cole believes this is that, as an added sweetener, his company enabled him to buy a house. His company did a share scheme, which meant that for every £1 he put in, they put 50p in. By the time he came to buy a house, it had increased by about six times in value. ‘It was a savings account that I never really touched.’ Cole didn’t have but nor did he need, the Bank of Mum and Dad; he was able to buy a house at 26 years old.

White Collar Deluge

The white-collar deluge is no longer a distant threat; it is already underway.

If we want to navigate it successfully, we must stop viewing learning as something that happens before work begins and start understanding it as something integral to the work itself.

This moment is not simply about responding to the rise of AI; it is an opportunity to fundamentally rethink the relationship between employer and employee. In a world where change is constant and linear careers are becoming increasingly rare, we need a new model.

That model requires companies to embed learning at every stage of working life, not just at the point of graduate entry. We cannot continue to treat people as disposable whenever the technology shifts. The old promise of a job for life may be gone. But it can and should be replaced with something more resilient and relevant: learning for life.

Because as AI continues to accelerate, the only sustainable future is one where human beings are not merely keeping up but actively moving forward, equipped with the skills to grow alongside it.

The companies that understand this will not only survive the coming transformation; they will become the kinds of organisations people truly want to work for.

Thanks for reading,

Eliza