Build #98 – AI and the workslop epidemic

Build #98 – AI and the workslop epidemic

I’ve been thinking about a particular kind of waste in business a lot recently.

Not the obvious kind – where someone clearly hasn’t done the work or has done it badly. That’s easy to spot and easy to fix.

No, I’m talking about something far more insidious: work that *looks* perfectly fine but is actually wholly worthless.

It’s work that appears plausible, sounds professional and uses all the right language. But when you really examine it, there’s nothing there. No insight. There’s no original thinking and no substance that actually moves anything forward.

We’ve got a name for it now: workslop.

Harvard Business Review defines this as content that “masquerades as good work but lacks the substance to meaningfully advance a given task.”

And this is destroying productivity, but not necessarily for the reasons you might expect.

The real problem with workslop isn’t that it’s obviously rubbish.

The real problem is that it’s plausibly adequate*

It passes a superficial inspection by looking like work has been done. It gets sent, read and responded to. And that means it creates more work downstream.

Here’s what makes it so dangerous for founders (and COOs): it’s invisible waste.

When someone doesn’t do their job, you know it immediately. When someone does it badly, you can usually see and feel the problems.

But when someone produces something that looks like good work? That seems to slip through too often.

It’s not just the illusion of progress that’s a problem. There’s another layer to this that’s even more of a problem for founders.

Workslop that gets created and then completely ignored.

As a COO I see this constantly with standard operating procedures (SOPs).

These look brilliant on screen and just the kind of documentation you’d be proud to show an investor during due diligence.

Except in the real world nobody’s following them.

Because nobody’s been trained on them and that’s because nobody actually knows they exist beyond the person who asked AI to create them.

I was working with a client last year who proudly showed me their ops handbook. It ran to forty-seven pages of beautifully formatted procedures covering everything from customer onboarding to quality control.

I’ve seen a lot of handbooks and playbooks, but my spidey sense was pinging as this one was too good.

I asked a few questions and learnt they’d created it in an afternoon using ChatGPT.

When I asked to watch their team actually onboard a customer, not a single step matched what was actually in the manual. They weren’t deliberately ignoring it – they literally didn’t know it existed.

The team member who had created it had done what they’d been asked to do. The founder felt good about having “sorted out their processes” and in reality no-one had made a difference to anything that matters.

And that’s not a one-off. I see it a lot.

This is workslop at its most wasteful. It’s not just meaningless, it’s actively misleading, creating the illusion that something has been done when actually, nothing has changed at all.

The gap between what’s written down and what’s actually happening becomes a kind of organisational fiction which nobody notices until something goes wrong.

Let’s get this straight…the work isn’t the document.

The work is the implementation, the training, the integration into daily operations and the often long-winded checking that people actually understand it.

But that’s hard. It takes time and that requires actual engagement with your team.

But of course it’s much easier to get AI to produce something that looks like you’ve done all that, file it away, and move on.

Here’s what I’ve realised: AI isn’t creating this problem. It’s just making it impossible to ignore. Businesses have always had a workslop problem with people going through the motions, producing outputs that look like work but lack genuine thought.

AI has just industrialised it, making it faster, easier and a lot more prevalent.

If your organisation is drowning in workslop, that’s a leadership failure.

AI doesn’t have judgement. It doesn’t understand what actually matters versus what just sounds good. It doesn’t know the difference between depth and the appearance of depth.

It produces output that looks like work because that’s what it’s designed to do.

The question is whether you can tell the difference and does your business operating system have the in-built checks and balances that weed out workslop.

Until leaders get serious about defining what good work actually means – what the right depth of thinking looks like and what shows a successful implementation has been completed – it’s only going to get worse.

So if I leave you with one thought today, it’s going to be that you need to start asking how wide the gap is between your documented reality and your actual one?

Would you even know?

> Read the article

Get more like this from Simon in your inbox

Build ditches startup hype to deliver raw, practical wisdom for founders about leading high growth businesses. Just straight talk from a fractional COO who’s seen every mistake, every shortcut and every hard truth founders have to face between start-up and scale-up.

“Great email. Thanks. It’s like you can see into the heart of my organisation. Sometimes what you send it so bloody timely it’s scary.”

“Timely and pertinent once again. I honestly don’t know how you do it!”

Join more than 1,000 founders learning how to scale without losing their minds, their team or their company culture.

Sign up now to get Build in your inbox.

About SIMON

I work as a fractional Chief Operating Officer (COO), consultant and advisor. I created the B3 framework® for company building and I also write a newsletter called Build for leaders who care about creating resilient and sustainable businesses.