Skip to content
Articles
ArticlesAI translation· 2 July 2026

The Dangerous Illusion of Competence

And why we're drowning in AI pseudo-expertise

by 5 min read

We live in fascinating but also absurd times. If you look at how digital society has evolved, the social web has established an entirely new metric for “competence”: reach and likes all too often trump solid expertise built on years of research and practice. That was a problem long before the current AI wave. But the wave brought a new iteration with it, which was already the subject of the very first post here on it's promp.td: the Simplicity Trap.

Episode 10 - The Turing effect and the Simplicity TrapDocIsInDaHouse

And plenty of people walked straight into that trap, which is how we ended up with this new wave of pseudo-expertise.

The analysis: a tsunami of AI pseudo-expertise

It used to be easier to spot a blowhard, because he lacked the vocabulary or the craft. His reach was small, and by the next day everyone had forgotten his drivel, because it wasn't written down anywhere and “shared”. Today? Today one of those AI chatbots/copilots/you-name-it that get pushed on us almost everywhere in the digital world encourages him to put his Dunning-Kruger syndrome on steroids.

Because one of the most striking features of this innovation, its extreme ease of use, is also its biggest curse. Anyone can now use a few simply worded prompts to churn out lofty, plausible-sounding texts, seemingly profound analyses or impressive graphics, which subversive algorithms then spread everywhere.

A real-world example:
Scroll through LinkedIn today (yet another tool urging everyone to use AI mindlessly). You will inevitably run into a flood of “AI evangelists & prompt gurus” trying to sell you, in a single post, the ultimate framework for implementing autonomous AI agents in large corporations. The fact that this person (if it is a person and not some automated bot) has never seen a critical change management project from the inside, or carried responsibility for a budget in the millions, disappears behind a seemingly perfect facade knitted together by algorithms.

Because the social media hype feeds on buzzwords and constant volume, a tsunami of “AI experts” is now rolling over us, all eager to explain with absolute certainty how the new world works.

The reality: expertise is not a prompt

We need to remind ourselves of one thing, again and again:

Real expertise and competence don't spontaneously emerge from the digital ether. They are the result of solid training combined with years of practical, often painful experience. Expertise is the distillate of talent, hard work and, crucially, mistakes you made yourself and learned from.

Let's put that in the context of the current AI hype:
The starting gun for this madness (ChatGPT) was fired a mere three years ago. The underlying technology is still immature, unreliable and changing at breakneck speed. It is simply impossible to have years of experience applying LLMs in a corporate context.

The people building real expertise here are mostly found in the engine rooms of companies right now. They experiment, they fail, they recalibrate. They know they don't have the absolute truth yet, precisely because the technology is in flux.

The result: maximum confusion

This gap explains the biggest problem I currently see in many consulting projects: the AI wave is driven less by rational facts than by marketing hype and pseudo-experts who have fallen for the Simplicity Trap. They confuse the ease of generating text with actually understanding the subject.

The result for business and society is confusion.
There's far too much noise and damn little real guidance.

The paradox: an important innovation in the wrong hands leads to ruin rather than salvation

This very confusion brings us to another phenomenon we can watch play out daily in a bewildering media landscape. The business news is full of contradictory headlines: on one side, incredible promises of salvation; on the other, more and more grim reports of spectacularly failed AI projects, millions burned on token budgets, or the imminent bursting of the AI bubble. Almost tragicomic are the regular reports of companies that, in a fit of pure AI activism, hastily lay off staff, only to sheepishly rehire the same people a few months later, because the promised “autonomous” AI can't get anything done without real human expertise after all.

This is the direct result of all this ubiquitous, toxic pseudo-expertise. When decision-makers seek advice from the loudest social media gurus, completely unrealistic expectations take hold. These “experts” sell seamless AI integration as a magic button you just have to press. Then bitter reality hits the projects: data protection problems, hallucinating models, outrageous API costs, and employees who corrupt the system with subversive “shadow IT”.

The paradox is this: AI is undeniably one of the most important innovations of our time. But in the hands of people who don't know the difference between a clever ChatGPT prompt and the safe integration of probabilistic systems into deterministic business processes, it becomes an expensive wrecking ball. The tools themselves are powerful, but many of these self-appointed architects haven't the faintest grasp of the building's structural engineering.

Skepticism, not FOMO

For many, it's this fear of being left behind, the classic Fear Of Missing Out, that blinds them to the warning signs and drives them straight into the arms of the loud saviours. But instead of panicking and chasing every trend, the exact opposite is called for.

If you've been reading this blog diligently, you mostly know it already: we're dealing with a highly complex, extremely resource-hungry and often unpredictable technology for which there are currently neither common standards nor final rules of the game. The sheer number of people slapping the AI label on themselves on LinkedIn or on big stages bears no relation to reality. The number of minds that really understand what's going on under the hood of these algorithms, how to assess them ethically and how to integrate them into the real world in a way that creates value, is frighteningly small.

Only a fraction of those who talk nonstop about AI today have the technical, process and strategic expertise it takes. And that's not something you can build up in a few months.

So if someone tries to sell you absolute certainty about AI, be suspicious. Anyone who claims to know exactly which way the wind will blow in five years, in a market whose rules change every month, is lying. At best, he's only kidding himself, caught in the Simplicity Trap.

At worst, he burns through your budget and lands you in court.


Live long and prosper 😉🖖


Soundtrack

Read on
Topic
Knowledge & Skills
12 PostsView all →
▶Listen