Between FOMO and stance: Marko Kalabota, the pragmatist of AI

There are conversations that begin as tech talk and end as a portrait of our times. My conversation with Marko Kalabota, CEO of Empolo, was one of those. We spoke about speed, about the fear of being left behind, about responsibility. And about why AI sometimes feels like an e-mountain bike: you finally make it up the hill, but you still have to pedal.

From eBay kid to AI collective

Kalabota is no brooding ivory-tower type. He is a doer. As a teenager he makes money off eBay hypes, later imports containers of furniture, delivers them himself, and finances his studies that way. Marketing becomes his terrain, social media his workbench. He knows machine learning from university projects, but in 2021 ChatGPT makes AI ready for the masses, and Marko conjures a business model out of it: Empolo as a collective, remote and hybrid, born out of the loft in Lochau. Marko Tovilo, co-founder of TOWA, is also on board at Empolo.

“Europe wants AI but often doesn’t know what for. So we bring the decision-makers into discovery workshops, sort ideas into short-term, medium-term, long-term, and identify the low-hanging fruit.”

That many companies already fail at their own infrastructure is no secret. Kalabota openly names many companies’ blind spot: data maturity. Without clean data, any AI architecture becomes a wish list. That is why Empolo first integrates into existing systems: from the hotel chatbot to the automated grant manager that asks questions, writes applications and defuses bureaucracy. The success can be measured against numbers.

Prototypes in days, not in months

What impresses me personally is his speed. “A prototype in a day,” he says. “That’s all possible now.” Of course: backend, roles, security come afterward. But the thesis stands: software development has become cheaper, creativity becomes a matter of timing.

“You used to need a designer and two developers. Today you orchestrate tools and models.”

That is not a disdain for craft, more the sober observation that orchestration is becoming a core competency.

The e-bike comparison and its limit

“AI is like an e-mountain bike. It carries you up, but it doesn’t ride on its own.”

I nod. The sentence explains why productivity can explode while responsibility still remains. It also explains why support automation at Klarna saves tens of millions and why robotics rebuilds entire value chains. The economic force is real. And it hits repeatable work first.

My follow-up: where does efficiency end, where does hollowing-out begin? Kalabota stays sober. Traditional companies that do not learn to use agents and APIs will be “swallowed.” Freelancers with tool competence lead teams of digital helpers. Atlas, Perplexity: all these tools become “assisted,” research and text take shape before the eyes of writers and service providers. Enthusiasm and unease sit so close together.

Creativity in the remix age

Kalabota is an artist. What unites us is a past as a creative, with more than 400 songs written and stored on the drive. He knows how much pop historically recycles. AI accelerates the principle. The difference: the capacity for fusion.

“Balkan with Korean patterns and Germanic brass: prompt it, mix it, make it sound unique.”

In Belgium, a grocery chain uses AI music and thereby saves on licenses. That is efficient, but it shifts the flows of value. I push back: where does the artistic core remain? Marko bets on authorship in the idea: the strategy, the message, the judgment of taste, all of that has to stay human. Fine-tuning yes, delegating the voice no.

The delicate zone: closeness, loneliness, edge cases

We touch on the uncomfortable: AI companions that say “love you, bro.” Adult bots that run on specialized models. Loneliness as a market. Kalabota names it without moralizing. His thesis: people seek genuine social friction, even when AI simulates closeness. The cycles of social platforms prove it:

1. social
2. then commercialized
3. then erosion
4. then a fresh start

At the same time, he warns against diagnostic promises and emotional shortcuts. This calls for ethics, transparency, oversight, and plain media literacy.

Europe between regulation and an action gap

“Europe is shooting itself in the knee with regulation.”

The US leads, Asia follows. He cites Singapore with its continuing-education subsidies, Oslo with ChatGPT Pro access for students, Albania with an AI ministry. Whether you agree with every example is secondary. His point is clear: the capacity to act beats the role of the perpetual skeptic.

Whoever in Vorarlberg, whoever in Austria fails to build the competencies loses in locational value. And that has consequences. “One keynote per company is not enough,” he says. It takes multi-week programs, real practice, a funding logic that empowers companies and employees. Here the regional government should provide even more help and support.

I agree, with one addition: rules are not a counter-design to innovation if they are pragmatic. Europe’s opportunity lies in secure infrastructures, verifiable provenance and data-sovereign architectures. Technology plus trust, that is a business model.

Responsibility that remains

Shortly before the close of our conversation, the topic of credibility comes into play. From the conversation with Hetfleisch I quote that the young generation grows up with the rule: everything is false until it is proven true. That can be healing, or cynical. That is why Kalabota, like Hetfleisch, relies on authentication and the attributability of content. I add: without media education, no watermark helps. The two together are protection.

What I take away from the conversation

• AI maturity does not begin with tools but with questions. What for? With which data? In which process?
• Orchestration becomes a key qualification: people who connect systems, clarify responsibilities, measure outcomes.
• Creativity stays human as long as we don’t outsource the idea. The style may be polished, the stance may not.
• Europe needs fewer buzzword strategies and more pilot projects that ease everyday work: grant managers, administrative co-pilots, care assistance, mid-market back office.
• Dignity is a design variable. Systems that monetize loneliness are technologically brilliant and socially poor. Here product policy decides the quality of the future.

Marko Kalabota sells the capacity to act. Empolo comes across like a workshop: fast in prototyping, open in consulting, clear in its claim to deliver value rather than show.

I leave the conversation with a simple image: yes, AI is an e-bike. It takes the gradient out of the route. But where we ride is up to us. And when we get off to truly take in the view, just as much. We may then see the summit after all, but for that we need our feet. And those can hurt.