This piece first appeared in German. The English version is a rewrite rather than a line-by-line translation, and the German original stays online in the archive: KI rechnet - wir denken.

Picture an analyst writing figures down by hand, cataloguing them in a ledger, and walking to a library or a records office for every piece of information the job requires. That was the work two generations ago, and nobody would go back to it.

The first wave of AI in professional work brought euphoria and the second brought a sense of being overwhelmed. What sits underneath both is a quieter change: processes run faster, decisions get more complex, and the rules shift without anybody announcing it. As the machine takes over more of the calculation, responsibility moves back onto the person using it, which is the part most discussions skip.

The Midjourney mentality

The term describes a specific failure: turning away from critical thinking inside the solving process. Take the output, ship the output. Errors appear that would never have survived an ordinary workflow, because nobody was checking at the point where checking used to be automatic.

Working with these tools without a long-term loss of quality requires the opposite posture. Engage critically with what the tool returns, and understand the problem deeply enough to know when the answer is wrong. Overcoming the habit of letting the machine work for you, and finding sound ways of working with it instead, is the actual goal.

None of this is new

Humanity has been through a process of automation before. Industrialization let people hand the production of goods to machines step by step, sparing muscle power. Textile production is the clearest case, where output rose to a degree that reorganized an entire industry within a few decades.

Those gains came with disadvantages for workers, because the occupations changed slowly and then completely. Work became innovation, optimization and expansion, carried on the shoulders of machines, and anyone unprepared for that shift was left behind.

What the value chain rewarded next

Three things gained importance in what we now call the value chain: operating the machines, building better machines, and optimizing the supply chain and the process around them. Automation of that kind is what allowed small workshops to grow into the international companies now found everywhere.

The process was considerably more complex than that summary and it played out over centuries. Today many of the largest companies, in finance above all, are not manufacturers in the classical sense. Even there the reliance on machines is total. Computers, laptops and phones are no longer separable from the work.

Looking at how much changed across those centuries raises an obvious question about the next few decades. Given the current pace, the consequences may not take centuries to arrive, which is the argument for engaging with the shift now and setting the direction deliberately.

Responsibility moves toward the person who checks

AI polarizes. Some see a job killer and others a productivity engine, and for professionals it mostly means the working world is changing differently from what the headlines suggest.

The historical parallels are unambiguous. Machines, computers, automation: every wave displaced routine and created more demanding work. AI belongs in that sequence. It takes the standard tasks and makes room for what machines cannot do, which is grasping context, making judgments, building narratives and creating trust.

The evidence is clearer than the panic suggests. So far there are very few AI-related layoffs, and a great deal of retraining, task redistribution and productivity gain. The opportunities are tangible and the risks are manageable, provided the change is accompanied by the right training and governance logic.

For professionals that means responsibility moves. The perfect spreadsheet counts for less, and the ability to translate between industries, datasets and cultures counts for more. Using AI as a sparring partner and not as a substitute for thinking is what makes someone effective.

Students are choosing inside a zone of imprecision

Being a student right now raises its own questions, and uncertainty brings opportunity with it. Which degree programs are dispensable and which are not? How do you make use of the current zone of imprecision, while it lasts? Which skills will be in demand?

Those questions are worth putting to people already doing the work. An academic discussion built on a student's knowledge base has its own interest, but often answers grounded in practice are more useful.

This was written jointly by Emil Ohligs and Sven von Bismarck. ThinkBeyondAI exists to supply orientation on exactly these questions: how AI changes value creation in consulting, law and finance, which competencies gain weight, and how the next generation of professionals learns to shape what comes rather than only operate the tools.

The link has been copied!