MIA
News & analysis

Why management loves AI and the shop floor doesn't — and why you're better off as a freelancer

An analysis of Glassdoor reviews shows that positive comments about AI have dropped from 81 percent in 2019 to 43 percent now. Executives are mostly positive, while frontline workers such as insurance claims handlers are almost entirely negative. I read it and thought straight away: this isn't about the technology at all. It's about who made the decision. And that's exactly where you, working for yourself, stand somewhere else.

Published 4 September 2026·3 min read·Source: The Decoder
Why management loves AI and the shop floor doesn't — and why you're better off as a freelancer

What the research actually says

The Decoder describes an analysis of reviews that employees leave about their own employer. In 2019, 81 percent of the remarks about AI were positive. Now it's 43 percent.

The split is the interesting part. Managers are mostly positive. Frontline staff, such as claims handlers at insurers, are almost entirely negative.

And the complaints go well beyond job loss. They're about forced adoption, about surveillance, and about productivity expectations that don't match the actual work.

The difference is who made the call

Put those two groups side by side and you see it immediately. Management chose it. The shop floor got handed it.

When you choose something yourself, a tool is something that makes your work lighter. When somebody else chooses it for you and attaches a higher tempo to it, it's something that happens to you. Same technology, different feeling. The difference sits in the decision, not in the software.

That's why I think this news is worth your time. You are both. You're the management and the shop floor of your own business. Nobody imposes anything on you. You decide yourself whether something makes your work lighter, and you can switch it off again just as easily. That's not a small luxury — it's precisely what those employees are missing.

Do the maths, let the numbers decide

This research doesn't change my advice. It only sharpens it: let the numbers decide, not somebody else's enthusiasm.

Take one task you currently dread. Your quotes, your inbox, your invoices, your messages. Write down three things:

  • How much time does that task cost you per month? Not by feel — keep count for a week.
  • What are you paying now in tools and outsourced work for that task? Your website package, your accounting software, the monthly bill for outsourced social media.
  • What is that time worth in the work you'd rather be doing? An hour spent on admin is an hour you don't invoice.

Then you have a number. Only after that can you answer whether an AI assistant is something for you. Without that number you're guessing, and then you're buying on feel.

Does the sum not work out in your favour? Then the answer is: don't do it. I say that about my own service too.

What you can do with this tomorrow

Concrete, and it takes about fifteen minutes.

Open your bank statements from the past month and mark everything that is a subscription or outsourced work. Website, hosting, accountant, planning tool, storage, marketing. Add it up. That amount times twelve is what your fixed digital work costs you this year.

Next to it, put the hours you spend yourself on the things you don't enjoy.

Only then do you look at whether automation changes anything about it. Maybe it does. Maybe it turns out you've already arranged things well and there's nothing to gain. Both are fine outcomes — at least you know.

If you'd rather not do that sum by hand, there's a calculator on my homepage that adds it up for you. Beyond that I need nothing from you. Questions are always welcome at info@mia-automation.com.

Frequently asked questions

I keep hearing employees dislike AI. Should I start with it at all as a freelancer?

That frustration is mostly about forced adoption, surveillance and expectations imposed from above. None of that applies to you. You choose yourself, and you can switch it off tomorrow. Start with one task that currently costs you money or time and see whether it helps. If it doesn't, you stop.

How do I know whether it actually gains me anything?

By doing the maths before you decide. Add up what you pay per month in tools and outsourced work, and count the hours you spend yourself. Put the cost of an AI worker for a small business next to it. If no clear difference comes out, the answer is no. I'd rather say that before you pay.

What does a virtual assistant cost compared to a fixed subscription?

A virtual assistant is billed per hour, so your costs rise with how busy you are. With me it's one subscription of 35 euros a month, cancellable monthly, no commission. Which works out better depends on how many hours it really involves in your case — so that's a sum you make first as well.