Transcript
Welcome to Iconoclast Insights, the podcast where we question the stories everyone else is selling as truth. I'm André Daus, and today we're talking about a narrative that's everywhere on LinkedIn but nowhere in reality—the idea that AI is replacing workers at scale. Let's dig into why this story persists and what it's actually hiding.
There's a narrative circulating in business circles that's becoming increasingly hard to ignore—not because it's true, but because it's repeated so often that it's starting to feel like truth. The story goes something like this: AI is replacing workers at scale, companies are becoming dramatically more efficient, and anyone who masters the right prompt will unlock unprecedented productivity.
Here's the problem: I can't find the evidence.
I read about this constantly on LinkedIn. Self-proclaimed CEOs of one-person operations posting about the AI revolution. Consultants selling courses on prompt engineering. Thought leaders predicting mass unemployment. But when I look at actual companies—the ones I work with, the ones in my network—I see something completely different.
I see organizations struggling to figure out where AI fits. I see teams experimenting cautiously. I see plenty of automation happening, but not in the dramatic, employee-replacement way the narrative suggests. So where is this story coming from? And more importantly—why are we believing it?
The Myth of Mass Replacement
Let's start with the most persistent claim: companies are replacing employees with AI at scale. I want to see the receipts. Not anecdotes. Not speculation. Actual examples.
Because here's what I'm seeing instead: companies are using AI to augment work, to handle repetitive tasks, to provide first drafts and analyses. That's not replacement—that's tool adoption. It's what happened with spreadsheets, with email, with project management software. The work changed. The roles evolved. But the wholesale elimination of positions? That's not happening in any systematic way I can observe.
Yet the narrative persists, and it's creating real anxiety. Workers worry their jobs are at risk. Executives feel pressure to "do something" with AI or appear behind the curve. Consultants sell solutions to problems that may not exist at the scale they're claiming.
This is narrative over data. And when narrative drives decision-making instead of data, we make bad choices.
One Prompt to Rule Them All
Then there's this almost magical thinking around prompts—the idea that somewhere out there exists the perfect prompt that will unlock AI's full potential and solve your problems. It's remarkably similar to the fantasy that somewhere there's a perfect productivity system, or a perfect organizational structure, or a perfect marketing formula.
It's nonsense.
Think about what a prompt actually is: it's an instruction set. It provides context and direction. But context is everything, and your context is unique. Your industry, your customers, your constraints, your objectives—these aren't universal. So why would there be a universal prompt?
The "perfect prompt" myth is dangerous because it suggests that success with AI is about finding the right words rather than understanding your actual problem. It reduces strategic thinking to a treasure hunt for magic phrases. It's intellectual laziness dressed up as innovation.
Here's the reality: working effectively with AI requires the same thing that working effectively with any tool requires—you need to understand what you're trying to accomplish, you need to understand the tool's capabilities and limitations, and you need to evaluate the output critically.
There's no shortcut to thinking.
The Accountability Gap
This brings me to perhaps the most troubling aspect of how AI is being deployed: the abandonment of accountability.
I see content on LinkedIn that's clearly AI-generated, published without any apparent review or consideration. I see companies implementing AI-generated recommendations without verification. I see the Deloitte case—247,000 euros refunded because AI-generated content containing fabricated sources and non-existent legal citations was submitted to a government client without proper quality control.
That's not an AI failure. That's a human failure.
The question everyone should be asking before publishing AI-generated content, implementing AI recommendations, or acting on AI analysis is simple: Would I put my name on this? Would I defend this? Do I actually believe this is correct?
If you're not asking these questions, you're not using AI—you're abdicating responsibility.
The European AI Act and similar regulations are trying to mandate transparency, but transparency without accountability is meaningless. If I tell you something was AI-generated but I haven't verified it, what good does that transparency do you? You still don't know if it's correct. You still don't know if I've done my job.
The tool didn't fail you. I failed you.
What AI Actually Does Well
Now, let's be clear about what AI can legitimately do, because dismissing it entirely is as foolish as the uncritical adoption I'm criticizing.
AI excels at pattern recognition, synthesis, and draft generation. It can analyze a legal database and provide preliminary assessments. It can summarize lengthy documents and highlight key points. It can generate first drafts that a skilled professional can then refine. It can identify trends in data that might take humans considerably longer to spot.
These are valuable capabilities. But notice what they all have in common: they require human judgment at the end of the process.
The AI provides analysis—you determine if it's relevant. The AI generates a draft—you decide if it's accurate. The AI identifies patterns—you evaluate if they're meaningful.
This is tool use, not replacement. And it's powerful precisely because it combines the AI's processing capabilities with human context, judgment, and accountability.
The Evolution of Tools
Here's the perspective I find most useful: AI is the latest iteration in a long evolution of information access tools.
Decades ago, if you needed specialized information, you went to a library. You spent hours searching through card catalogs and indexes. You read through multiple sources to find what you needed. The information was there, but access was slow and labor-intensive.
Then came CD-ROM encyclopedias. Suddenly you could search thousands of articles from your desk. The information was the same, but access was faster.
Then came Wikipedia and search engines. Now you could access millions of sources in seconds. Same fundamental task—finding and evaluating information—but dramatically faster execution.
AI is the next step in this progression. It doesn't just find information—it synthesizes it, analyzes it, and presents it in formats tailored to your query. But the fundamental challenge hasn't changed: you still need to evaluate what you're getting. You still need to determine if it's accurate, relevant, and useful for your specific context.
The tool keeps getting better. The human responsibility remains constant.
The Real Risk
Here's what concerns me most about the current AI discourse: we're so focused on whether AI will replace us that we're missing the actual risk.
The risk isn't that AI will do our jobs. The risk is that we'll let AI define what our jobs are.
If we start letting AI tools tell us what problems we have, what questions we should ask, what solutions we should pursue—if we outsource strategic thinking to algorithms—then yes, we make ourselves redundant. Not because the AI is replacing us, but because we've stopped doing the work that actually matters.
Strategic thinking isn't about finding answers. It's about asking better questions. It's about understanding context that isn't obvious. It's about challenging assumptions that everyone else accepts. It's about seeing connections that don't show up in the data.
These capabilities are inherently human. They require judgment developed through experience, intuition refined through failure, and the ability to hold multiple contradictory ideas simultaneously while working toward synthesis.
AI can't do that. It can process information, but it can't understand meaning. It can identify patterns, but it can't determine significance. It can generate options, but it can't make decisions that account for all the messy, unmeasurable factors that drive real-world outcomes.
What This Means for Strategy
If you're a business leader trying to figure out how AI fits into your organization, here's my advice: stop asking "How can AI replace X?" and start asking "Where are we doing low-value work that AI could handle, freeing our people to focus on high-value thinking?"
The question isn't about replacement. It's about allocation of human attention.
Where are your talented people spending time on tasks that don't actually require their judgment? Where are routine processes consuming resources that could be redirected to strategic challenges? Where could AI handle the groundwork so humans can focus on the interpretation?
These are the right questions. They lead to productivity gains without the fantasy of wholesale replacement. They improve efficiency while actually increasing the value of human contribution rather than diminishing it.
The Uncomfortable Truth
The AI replacement narrative is appealing because it's simple. It gives us a clear story about the future. It creates urgency and justifies action. But simple narratives are usually wrong, and this one is particularly dangerous because it distracts from real opportunities while creating imaginary threats.
The uncomfortable truth is this: most organizations don't need AI to replace people. They need AI to expose the inefficiencies, unclear processes, and strategic gaps that were already there. AI doesn't create these problems—it just makes them more visible by highlighting where human judgment is absent, where quality control is lacking, where strategy is unclear.
That visibility is uncomfortable. So instead of addressing the underlying issues, we look for simple explanations: "AI is overrated" or "AI will replace everyone." Both are wrong. Both let us avoid the real work.
The real work is examining our own assumptions, clarifying our strategies, and determining where human judgment actually adds value versus where we're just going through motions that could be automated.
That's harder than blaming or praising AI. But it's what actually matters.
Moving Forward
So where does this leave us?
AI is a powerful tool that's becoming increasingly capable. It will continue to change how work gets done. But the narrative about mass replacement is—at least currently—more myth than reality. And the fantasy of the perfect prompt is a distraction from the real challenge of strategic thinking.
What we need is not better prompts or more AI tools. What we need is clearer thinking about what problems we're actually trying to solve, better judgment about when tools are appropriate, and stronger accountability for the outputs we produce—whether AI-assisted or not.
The question isn't whether AI will replace you. The question is whether you're willing to do the thinking that AI can't.
Because that's where the real value lies. That's where human beings remain irreplaceable. And that's where focusing our energy will actually move organizations forward rather than just creating the appearance of progress.
Your thinking creates your limits. And right now, our thinking about AI is limiting us more than the technology itself ever could.
