Everything Has a Beginning

The jobs crisis will not begin with mass redundancies. It will begin with vacancies that quietly disappear.

Everything has a beginning.

A financial crisis begins with a missed payment. A recession begins with a cancelled order. And a jobs crisis does not necessarily begin with millions of people being handed redundancy notices on the same morning.

It can begin far more quietly—with jobs that are simply never advertised.

The latest UK labour-market figures may prove to be one of those moments. The number of vacancies has fallen to 702,000, its lowest level for five years. Strip out the exceptional Covid period and vacancies have not been this low since 2014.

At first glance, that may look like another disappointing but unremarkable economic statistic. It is not. It may be an early warning that the relationship between businesses and human labour is beginning to change.

The First Stage Is Not Redundancy

When people imagine automation taking jobs, they tend to picture something dramatic: robots entering factories, offices being emptied or thousands of workers replaced overnight.

The reality is likely to be much less theatrical.

A business faced with higher wages, increased employment taxes, weak demand and stubborn inflation does not necessarily dismiss ten people and replace them with machines. Instead, when somebody leaves, the company pauses before replacing them.

Can their duties be divided among the remaining staff? Can software absorb the administration? Can artificial intelligence draft the documents, answer the routine enquiries, analyse the information or manage the diary?

Increasingly, the answer will be yes.

No dismissal is recorded. No factory gates are closed. There is no announcement saying that artificial intelligence has taken another job. There is simply one fewer vacancy.

Multiply that decision across tens of thousands of businesses and the labour market can undergo a profound change without producing a single defining moment.

Businesses Are Already Under Pressure

The Office for National Statistics says smaller companies continue to report that increased labour costs are affecting recruitment decisions. Payrolled employee numbers have generally been falling for two years, while the provisional August estimate was 145,000 lower than a year earlier.

Retail and hospitality—two of the most important entry points into employment—have experienced particularly noticeable falls.

For now, the immediate explanation is cost. But cost pressure is precisely what encourages automation.

Artificial intelligence does not have to become perfect before it changes employment. It only has to become cheaper, quicker or sufficiently capable to make the next hire unnecessary.

That is an important distinction. The threat is not simply that AI will replace existing workers. It is that companies may gradually stop creating the positions through which young people become experienced workers in the first place.

Where Does the Next Generation Begin?

Every experienced accountant was once a junior accountant. Every senior analyst began with routine research. Every accomplished solicitor once reviewed basic documents. Every sales director started by making calls, learning objections and making mistakes.

Entry-level work has never been glamorous, but it has always served a vital purpose: it converts education into experience.

AI is particularly capable of absorbing many of the tasks traditionally given to people at the beginning of their careers. If those first rungs disappear, we do not merely lose jobs today. We weaken the mechanism by which expertise is created for tomorrow.

Young people may be told to gain experience while businesses simultaneously remove the jobs that provide it. Universities could continue producing graduates for roles that exist in ever-smaller numbers. Competition intensifies, wages weaken and adulthood is delayed—not through a lack of ambition, but through a lack of opportunity.

The consequences then spread far beyond employment statistics.

The Economic Feedback Loop

People without secure employment spend less. They delay moving out, buying homes, starting families and building pensions. Lower household spending reduces business revenue, which gives companies another reason not to recruit.

At the same time, government receives less in income tax and National Insurance while spending more on welfare and support. Housing demand weakens. University applications become harder to justify. An ageing population becomes increasingly dependent upon a smaller base of productive taxpayers.

This creates a dangerous feedback loop:

Higher employment costs lead to fewer hires. Fewer hires encourage greater use of AI. Greater automation suppresses entry-level opportunity. Weaker household incomes reduce demand. Lower demand causes businesses to recruit even less.

Technology may increase output, but an economy cannot thrive indefinitely if a growing share of its population lacks the income needed to purchase that output.

The Productivity Promise Has a Flaw

We are frequently told that artificial intelligence will reduce costs, increase productivity and therefore make goods and services cheaper.

Perhaps it will in some competitive markets. But there is no law requiring a company to pass its savings to the customer.

Businesses automate primarily to protect margins, increase profits or outperform competitors—not as an act of charity. If a company can produce the same output with fewer employees, the immediate financial reward belongs to the owners of the technology and capital.

Unless those gains are distributed more widely, the result could be an economy producing more with fewer people while concentrating a greater share of the rewards among fewer companies and shareholders.

The goods may exist. The productivity may exist. The profits may exist. But where does the mass purchasing power come from?

That may become one of the defining economic questions of the next decade.

Inflation Makes the Transition More Dangerous

This change is not happening in a stable economic environment.

The UK unemployment rate remains at 4.9%, but inflationary pressure has not disappeared. The Bank of England now faces the deeply uncomfortable possibility of raising interest rates to restrain prices while private-sector employment is already weakening.

That combination matters. If employment falls during a period of low inflation, policymakers can cut rates and stimulate demand. If jobs weaken while inflation remains elevated, their choices become far more limited.

Households could face the worst of both worlds: less employment security alongside higher food, energy, housing and borrowing costs.

That is not the painless transition to an AI-powered future we have been promised. It is the foundation of a potentially prolonged stagflationary squeeze.

It Will Not Arrive Wearing a Name Badge

We should be careful not to claim that every missing vacancy has been destroyed by artificial intelligence. The current evidence does not prove that. Higher wages, taxation, interest rates and weak business confidence are already sufficient to explain much of today’s slowdown.

But waiting for a statistic labelled “jobs lost to AI” would be a serious mistake.

Economic transformations rarely announce themselves clearly at the beginning. They appear first as disconnected pressures: a graduate scheme reduced here, an administrative position left vacant there, a team expected to deliver more without replacing somebody who has left.

Only later does the pattern become undeniable.

The most revealing part of the latest figures is therefore not that unemployment remained unchanged. It is that small firms are scaling back new hires.

That is how structural unemployment can begin—not with mass redundancies, but with mass non-recruitment.

Prepare Before the Consequences Become Obvious

Periods of profound technological and economic change create winners, but they also create instability. Labour markets weaken. Tax revenues come under pressure. Governments borrow more. Political demands for intervention grow. Currency creation and redistribution become increasingly tempting responses to a problem that conventional economic policy may struggle to solve.

For investors, this is not an argument against technology. AI may prove to be the most productive invention of modern history. It is an argument for recognising that enormous productivity gains can still arrive alongside social disruption, fiscal strain and monetary debasement.

That is precisely why diversification matters.

Physical gold cannot solve the employment crisis and it cannot determine how the rewards of automation are shared. What it can provide is ownership of a finite asset outside the employment market, outside corporate promises and outside the banking system—one that has historically been used to preserve purchasing power through periods of economic and monetary upheaval.

Everything has a beginning.

Perhaps, years from now, we will look back at collapsing graduate opportunities and decade-low vacancies and recognise that the AI employment crisis had already started—not with a bang, but with a job advert that was never placed.

About the Author

Matthew Jones is Co-founder and Precious Metals Analyst at Britannia Bullion and a contributor to Kitco News. His work focuses on gold, monetary risk, economic change and long-term wealth preservation. Britannia Bullion is a UK-based precious-metals specialist helping private clients protect, preserve and pass on wealth through the ownership of physical gold and silver. The company combines market education and ongoing analysis with a highly personal, consultative service supporting clients through purchasing, insured delivery, independent secure storage and its guaranteed buy-back service.