I Don't Have Time for Your AI Slop

A personal perspective on AI, responsibility, and the value of our time.

I'm tired of cheaply produced AI content.

And when I talk to other executives, managers, and entrepreneurs, I realise I'm not alone.

Don't get me wrong. I'm excited about the possibilities of artificial intelligence. I'm investing in this technology myself, working with it every day, and I'm convinced it has already fundamentally changed the way we work – and will continue to do so.

Why else would we have founded Arclumina AG, a company dedicated to AI? To me, AI is the steam engine of our time – the driving force behind a new industrial revolution.

Ten Pages of Text. Zero Substance.

Sound familiar? You ask someone for an analysis, an assessment, a decision-making document, or a presentation.

And what comes back? A ten-page document. Perfectly structured, professionally written, complete with an executive summary, strategic priorities, opportunities and risks, and – of course – a final recommendation.

At first glance, everything looks great. But by the second page, you realise that someone has barely engaged with the actual problem. Instead, they've asked an AI to produce as much convincing-sounding text as possible.

The result? AI slop.

We Can All Spot It by Now

Typical AI slop has a surprising number of things in common:

• Statements that sound important but don't actually say anything.
• Endless lists of opportunities, risks, and recommendations, with no clear priorities.
• Buzzwords like "holistic", "strategic", "sustainable", and "future-oriented" scattered throughout every other paragraph.
• The same arguments repeated over and over, just worded differently.
• Recommendations so generic they could apply to practically any company.
• Lots of structure, lots of text, but no discernible personal opinion.

And my personal favourite: three paragraphs explaining why something matters without ever telling us what we should actually do.

The problem isn't that these texts were generated by AI. The problem is that apparently nobody bothered to critically review the output.

My Time Is Too Valuable for This

When someone sends me this kind of work, it usually goes straight back to the sender.

I'm not willing to spend twenty minutes of my time reading something that someone else generated in twenty seconds. Why should I do the thinking that the sender couldn't be bothered to do?

That's neither efficient nor particularly respectful of other people's time.

AI should make our work better. Not simply longer.

AI Is Already Changing Our Language. But What About Our Thinking?

The influence of AI on our language is now measurable. A recent analysis by ZHAW, reported by the NZZ (well worth reading!), shows that typical AI expressions have become significantly more common in Swiss German-language media since the rise of ChatGPT. The German word "Herzstück" ("centrepiece"), for example, is now used about twice as frequently as before. Of course, that doesn't automatically mean all these texts were written by AI. Apparently, we're also adopting linguistic patterns we're repeatedly exposed to.

And that's what gets me thinking. Because if AI increasingly influences how we express ourselves, it may also influence how we argue, describe problems, and draw conclusions.

«If AI shapes our language, sooner or later it will shape our content too.»

A 2025 study by Carnegie Mellon University and Microsoft Research (a bit lengthy, but also worth reading!) suggests that this concern isn't entirely unfounded. The researchers surveyed just over 300 knowledge workers about how they use AI. One particularly interesting finding: the more participants trusted AI's responses, the less critical thinking they reported applying to those tasks. Conversely, greater confidence in their own abilities was associated with more critical thinking.

Of course, this doesn't prove that AI is making us forget how to think. But it's certainly a warning sign. What happens when we stop developing our own arguments and increasingly adopt those generated by a machine? When we stop questioning statements simply because they sound plausible and professionally written?

We risk not only producing more interchangeable content, but eventually thinking in more interchangeable ways too.

Back to Basics, Please!

We need to relearn how to focus on what really matters.

That doesn't mean using less AI. Quite the opposite. We need to learn how to use it better. Learn how to write good prompts. Be clear about what you actually want to know. Challenge the AI. Demand critical arguments instead of polished generalities.

And above all: Read what the machine has written. Question it. Cut it in half. And then perhaps cut it in half again. Ask yourself whether you truly understand the statements, whether they're relevant, and whether you actually stand behind them.

A good analysis doesn't need to be ten pages long. Sometimes three precise sentences are worth more than thirty perfectly polished paragraphs.

Thinking Is Still Our Responsibility

We risk using AI not only to become more productive, but also to produce more meaningless content.

Just because a machine can express our thoughts more elegantly doesn't automatically make those thoughts any smarter. And just because it can turn a simple idea into ten pages doesn't mean we should send those ten pages to someone else.

Use AI to think better. Not to skip the thinking.

And please: Stop writing nonsense just because a machine makes it sound good.