AI Won’t Cause Unemployment

There’s a lot of chatter out there about AI and how it’s going to make us all jobless. If you scroll through social media (something I can’t recommend from personal experience), you’ll see endless takes: that AI will replace writers, data analysts, programmers, executive assistants… basically everyone who works with a computer. So, is it all just scaremongering?

As I work through my own AI automation projects, I’ve given a lot of thought to the idea that AI will cause unemployment. Spoiler: I don’t think AI will cause unemployment, but it will fundamentally change how we work. So, stop freaking out (at least about the wrong thing).

Yes, AI Will Make Some Jobs Redundant

The big idea that’s spooking people: AI can do a lot of “data shuffling” tasks in a fraction of the time it takes a human. For example, summarizing spreadsheets, generating lengthy reports, rewriting content for different audiences, scheduling tasks — these are typically the domain of data analysts, writers, or assistants. This is why I say that middleware jobs will cease to exist.

Data from Oxford Economics found that by 2030, up to 20 million manufacturing jobs worldwide could be replaced by robots alone, and that’s before we talk about LLMs and other forms of AI.

If all you do is move information from a spreadsheet to a slide deck, or from one database to another, you’ll be easily replaced by a sufficiently advanced automated tool.

In short, “middleware jobs” — the layers of people responsible for translating, copying, processing, or summarizing information — are going to either disappear or massively transform in nature.

But to stop there would be foolish. Because what’s more likely to happen, based on past experience, is that new forms of work will emerge, existing jobs will expand in scope for those able to keep up with the tech, and the growth rate of the economy will accelerate. AI isn’t just about replacing tasks; it’s also about creating capacity.

The “Efficiency Boom”

When we talk about AI, we often forget it’s a tool, not a magic genie. It won’t sit in a corner and imagine how to do your job for you (without being pointed in the right direction!). Businesses still need people to guide AI, verify the output, and make decisions — many of which still involve human empathy and strategic judgment.

• Writers: Tools like ChatGPT or Claude can draft content, but humans need to provide context, brand voice, and strategic direction. It means one writer can do the work of ten, but there’s still a writer (or at least an editor) in charge.
• Data Analysts: AI can ingest data and produce analyses with lightning speed. But an analyst still needs to figure out the right questions to ask, interpret the results, and communicate them effectively in the context of the business strategy.
• Programmers: AI-based code assistants can write or optimize code, yet humans are needed to design architecture, maintain code integrity, and ensure everything aligns with product and company goals.
• Executive Assistants: AI can schedule your meetings and draft your emails, but you still need a human who understands office politics, company culture, and how to gracefully handle a crisis in scheduling.

In all these scenarios, AI doesn’t completely remove the person — it amplifies their ability. A single person can do more, and do it faster, than they could before. Far from mass layoffs, companies are more likely to invest in people who can harness these tools to produce 10x the output.

If you (or your business) can produce more output, you can often serve more customers, which brings in more revenue. That means more jobs in associated fields that don’t directly involve “middleware” tasks. The pie isn’t a fixed size.

All that 10x’d output doesn’t serve the same demand; many other industries also benefit from the same growth in capacity. So what happens is that the economy grows as a whole.

The Industrial Revolution Parallel

The classic historical precedent for automation angst is the Industrial Revolution. Steam-powered machines replaced many manual labor jobs. Textile mills equipped with power looms reduced the need for skilled weavers. People (quite understandably) felt threatened, as many skill sets became obsolete.

But what actually happened?

1. Productivity Soared — Factories churned out far more goods.
2. Demand Increased — Because goods were cheaper to produce, people could afford more, driving economic growth.
3. New Jobs Emerged — From machinists to distribution managers, entirely new categories of work popped up.

Human history is filled with these transformations: mechanized agriculture drastically reduced the number of farm workers but spawned roles in other industries. The net effect was an expanded economy and, generally, more employment options.

However, it’s also true that the rich and powerful benefited the most. Workers did remain employed, but the fortunes accrued fastest to those who owned the factories and machinery. The lesson? Yes, automation can make everyone more productive — but it can also exacerbate wealth inequality.

Inequality Will Rise

Here’s the reality check: If you own or invest in the AI tools that exponentially boost productivity, you stand to gain a lot. If you’re “just an employee,” you also benefit (your skills are more leveraged, and your wages might rise due to higher output).

But usually, the greatest chunk of value accrues to those at the top. We saw this with the industrialists of yesteryear, and we’re seeing it again with AI entrepreneurs and investors today.

Thus, while I don’t think AI will cause unemployment in the long term, it will intensify inequalities unless we build systems (e.g. progressive taxation, robust upskilling programs, etc.) to counterbalance those forces.

Don’t Bet on Universal Basic Income

Many people wonder if governments will consider giving everyone a UBI (Universal Basic Income) once AI does everything. The idea is: “If most jobs become redundant, then we might need a global allowance so people can live.”

But mass unemployment is unlikely. That means we probably won’t have the political or social impetus for UBI. Historically, governments have only ever provided drastic safety nets (e.g. war pensions, unemployment benefits) in response to large-scale crises.

If the job market remains relatively robust, albeit changed, don’t expect a crisis large enough to incentivize a universal safety net.

UBI might remain an interesting experiment or be implemented in a small scale (some localities have tested it), but widespread adoption would require a level of unemployment that we likely won’t see. People will be employed — just differently, and with increasingly efficient tools at their disposal.

How to Thrive in the AI Age

Given that we won’t all be mass unemployed, how do we prepare ourselves to thrive? A few ideas:

1. Get ahead of the curve: AI is evolving constantly. Regardless of your field — even if it’s not tech — learn how the tools work and how they can help you. Get ahead, as I decided to, by building the tools. If you’re in sales, for example, look into AI-driven customer analytics or automated outreach.
2. 10x your putput: If you don’t want to build the tools, then you can at least use them. Step-up your output by getting tools to do the menial work.
3. Help others: There’s always room for people to help others who can’t keep up. Tutor old folks on not just using computers but on using these tools — there’s a gold mine of opportunities there.

The Bottom Line

AI isn’t going to cause widespread unemployment. It will disrupt people who only do repetitive “middleware” tasks — but history suggests the economy will expand around new efficiencies. Those who adapt and learn new tools will become more productive, and new industries will be born.

But don’t forget that the spoils will (yet again) go to those who own the tools. If our track record is any guide, that will mean growing inequality, even if we’re all collectively richer.

People and governments need to think about how to redistribute some of that wealth so everyone can benefit, not just top-tier moguls. As much as we all want a go at being a mogul…

In the end, more work will shift to direct human interaction or creative endeavours — exactly the kind of stuff we actually wanted to do as kids. Embrace that future, get over the doom and gloom, and figure out how to make the most of the transformation.