AI isn’t taking our jobs, it’s just giving us more work

4 hours ago 3

Opinion

October 23, 2025 — 7.55pm

This week, a documentary aired in Britain about automation and the job market. Channel 4’s Will AI Take My Job? featured a television reporter walking the streets of London as she contemplated which jobs might be at risk in the near future.

Dressed smartly in a crisp white shirt and tailored jacket, she began: “Call centre workers, customer service agents. Maybe even TV presenters like me,” she paused, “because I’m not real.”

Channel4’s AI-generated presenter Aisha Gaban.

The twist of the show was that the very realistic host who had been taking the audience through the journey was AI-generated. Her voice and likeness were all figments of a robot’s imagination, designed to push our uncomfortable human buttons about where all of this is heading.

One of the most consistent messages we’ve heard is that no job is safe in AI’s eyes because it’s going to either do the work for us or make us so efficient that there will be nothing left for humans to do.

But what if instead of making us more efficient, AI might actually be doing the opposite?

That’s one of the conclusions from new research from Stanford that found many employees are now using AI tools to create low-effort and low-quality output that ends up creating more work for their co-workers. Riffing on the existing idea of AI slop, researchers have named this “workslop”.

It seems that each day a new AI tool is released to seemingly help us be more efficient, each product making it easier for workers to produce passable content at the click of a button. This means we’re beginning to drown in lengthy reports written with a few words of a prompt, or slides that look thoughtfully put together.

But underneath this shiny surface is the sad reality that some of the output is actually unhelpful, misleading, incomplete or not properly understood by the person creating it.

According to the Stanford research, of more than a thousand US workers, 40 per cent reported they had received some version of workslop in the past month, mainly from their peers but also handed down from managers to direct reports.

I recently heard a story about two people doing business, each using AI in an attempt to be more efficient. The first person wrote a few simple bullet points into ChatGPT to create a long, professional-sounding document to send to the potential business partner. When the other person received it, it was so wordy that they asked ChatGPT to summarise it back into a few simple bullet-points.

We’re now adding a real cost of trying to battle through the noise to find out what the human behind it really means.

This is a prime example of workslop, in which AI is used unnecessarily to create more work at both ends, instead of both parties just exchanging simple bullet points in the first place.

There are real-world consequences of outsourcing thinking to AI, just ask Deloitte Australia, which will take years living down its recent scandal of using AI to partially create an error-riddled $440,000 report for the Department of Employment and Workplace Relations.

We may fool ourselves into thinking that everything we do with AI is making us more efficient, but we’re now adding a real cost of trying to battle through the noise to find out what the human behind it really means. This process shifts the mental load from the creator to the receiver, and that’s extremely unhelpful.

It also makes people think less of anyone who sends them low-effort work, which is the opposite of what they intended.

So, is AI going to take your job? I used to always lean towards yes with this question, but unless we can get workslop under control before it becomes an epidemic, I’m now no longer certain.

Tim Duggan is author of Work Backwards: The Revolutionary Method to Work Smarter and Live Better. He writes a regular newsletter at timduggan.substack.com

Get workplace news, advice and perspectives to help make your job work for you. Sign up for our weekly Thank God it’s Monday newsletter.