The coming AI jobs-pocalypse

Every major technological revolution has triggered waves of anxiety about job losses, social disruption, and economic dislocation. From the mechanisation of agriculture to the rise of computing, each era has felt uniquely destabilising to those living through it. And yet, over time, new forms of work have emerged, often in ways that were difficult to predict at the outset.

The AI revolution is no different in its emotional texture. What may be different is its speed, scope, and the kinds of cognitive work it reshapes.

To understand what might come next, it helps to look backward.

The agricultural to industrial shift

Before the industrial revolution, the majority of people worked in agriculture. In countries like the United States and the UK, upwards of 60 to 80% of the workforce was engaged in farming or farm-adjacent labour.

Mechanisation changed that. Tractors, harvesters, and industrial processes dramatically reduced the need for human labour in agriculture. By the mid-20th century, agricultural employment in advanced economies had dropped to single digits.

Jobs lost:

Manual farm labourers

Seasonal agricultural workers

Rural craft and subsistence roles

Jobs created:

Factory workers

Mechanical engineers

Transport and logistics workers

Urban service roles such as retail, hospitality, and administration

What’s striking was not just the shift in jobs, but the shift in where and how people lived. Entire populations moved from rural to urban environments. Work became more structured, clock-driven, and centralised. This revolution changed our relationship to work, seasons, time and nature. No biggie!

Electrification and mass production

The next wave, electrification and assembly-line production, further transformed work. It didn’t just eliminate jobs, it reorganised them.

Jobs lost:

Skilled artisanal manufacturing roles such as hand-weaving and bespoke production

Small-scale local production

Jobs created:

Assembly line workers

Industrial managers

Electrical engineers and technicians

Consumer economy roles such as marketing, sales, and distribution

It also shifted the personal agency that workers had in their work from high levels of agency in the skilled artisinal work to low levels of agency in the factory assembly line work. This period also saw the rise of the modern corporation and managerial class. Coordination, planning, and oversight became distinct forms of labour. The rise of the managerial class came about because we needed humans to help coordinate labour.

The digital and computing revolution

From the late 20th century onward, computing began to automate routine cognitive and administrative tasks.

Jobs lost:

Typists and stenographers

Filing clerks

Switchboard operators

Some middle-management coordination roles

Jobs created:

Software developers

IT support and systems administrators

Data analysts

Digital marketers

Entirely new industries such as e-commerce, social media, and cybersecurity

This shift is different because it is the beginning of an encroachment on white collar work. It also introduced the idea that information itself could be processed, stored, and scaled independently of human labour.

A pattern emerges

Across these revolutions, a few patterns repeat:

Technology tends to eliminate tasks, not entire jobs, at least initially

New jobs often emerge at the intersection of the technology and human needs, including managing, interpreting, selling, or maintaining the technology

The transition period is uneven and often painful for displaced workers

The new jobs are rarely obvious at the outset

Most importantly, each wave has moved human labour up the stack, away from physical effort and routine tasks, and toward coordination, creativity, and judgment.

The AI revolution: what’s different?

AI has changed the cyber threat landscape, amplifying risks while reshaping defences. AI is already demonstrating the ability to perform tasks that were previously considered uniquely human, including writing, coding, analysing, designing, and even aspects of decision-making.

This creates a qualitatively different kind of disruption.

Jobs at risk are not just manual or routine. They include:

Entry-level knowledge work such as junior analysts, paralegals, and copywriters

Administrative coordination roles

Basic programming and software development tasks

Customer service and support functions

At the same time, new roles are emerging:

AI system trainers and evaluators

Prompt engineers and workflow designers

AI governance, risk, and compliance specialists

Human-AI interaction designers

Data curators and synthetic data engineers

More interestingly, AI is not just creating new job categories, it is reshaping existing ones.

For example, a lawyer with AI assistance can do the work of several junior associates. A consultant can produce analysis faster and at greater scale. A solo creator can operate like a small media company.

This suggests that the impact of AI may be less about wholesale job replacement and more about:

Workforce compression, where fewer people do the same work

Skill amplification, where high performers become significantly more productive

Role hybridisation, where jobs blend technical, creative, and strategic skills

The uncomfortable middle

If history is a guide, the biggest challenge is not the end state, it is the transition.

I've seen this play out in previous technology revolutions. For example, when I worked on a project that replaced the typing pool with automated letter generation in an insurance business. Workers displaced by agricultural mechanisation did not seamlessly become factory workers. Nor did typists automatically become software developers. These transitions involved generational shifts, retraining, and often significant social upheaval. The same is likely to be true for AI.

We may see:

Increased inequality between those who can effectively use AI and those who cannot

A hollowing out of entry-level pathways into professions (unless we actively work to ensure that these pathways evolve for the AI age)

Pressure on education and training systems to adapt rapidly

A rethinking of what constitutes valuable human work

What remains uniquely human?

Every technological revolution has forced a re-evaluation of human value. As machines take over routine physical and cognitive tasks, the remaining areas of human advantage tend to cluster around:

Judgment under uncertainty

Ethical reasoning and accountability

Complex interpersonal interaction

Creativity and original synthesis

Contextual understanding across domains

These are not new capabilities, but they are becoming more economically central.

Looking forward

It is tempting to ask whether AI will create more jobs than it destroys, as if history guarantees a reassuring answer. History suggests something more nuanced.

Yes, new jobs will emerge. But they may not emerge quickly enough, or in the right places, or for the same people who lose their jobs. And they may require fundamentally different skills. Governments will probably need to step in and reimagine how they support people through this transitional period.

The real question is not whether jobs will exist, but what kinds of work will be available, to whom, and on what terms.

In that sense, the AI revolution is not just a technological shift. It is a societal and a governance challenge, an economic restructuring, and a societal choice.

One thing is certain, the pace of change will not slow down, and we will all have to become lifelong learners just to keep up.

And, like every revolution before it, the outcomes will depend less on the technology itself and more on how we as nations and societies choose to deploy, regulate, and adapt to it.