AI needs data. Starting something new still takes human courage, says Vladimír Bezděk, advisor to the Czech president and AVANT Investment Company chairman

August 6, 2025

Vladimír Bezděk, economic advisor to the Czech president Petr Pavel and Chairman of the Board at AVANT Investment Company, discusses the growing role of AI in the financial sector and beyond. From comparing investment fund statutes to supporting asset valuations, artificial intelligence is becoming a reliable assistant—, yet it still raises complex questions.

• What decisions can AI handle better than humans?
• Where does it fall short?
• And what will be left for junior professionals once AI masters routine knowledge work?

Read the podcast as an interview

(The interview was shortened and edited using ChatGPT)

Ivana Karhanová: How are you personally — or at AVANT — using artificial intelligence?

Vladimír Bezděk: Personally, I’d say I’m at level 4 out of 5. I’m learning, like many others. There are blind alleys, and I’ve learned that you can’t always rely on AI for facts. At AVANT, we’re experimenting mainly in areas that involve large volumes of text.
We manage qualified investor funds, and each fund has a statute. If a fund has existed for 5 or 10 years, its statute may have changed multiple times. But when dealing with investors, we must still honour the conditions that applied when they joined. This means checking historical versions of long, detailed documents — a process prone to human error. That’s where AI could help, and where we’re currently testing it.

Ivana Karhanová: And beyond this use case —, do you see other real applications?

Vladimír Bezděk: Yes, several. One potential area is preparing valuation drafts for certain types of assets, especially if AI can learn from past appraisals. These drafts would still need expert review, but even that would save time.

Ivana Karhanová: Because the valuation process usually follows a consistent framework…

Vladimír Bezděk: Exactly. And that structure makes it teachable.

Ivana Karhanová: Are you exploring other types of outputs beyond documents and text-based decisions?

Vladimír Bezděk: Some of the outputs are numbers or decisions. — for example, determining investor entitlements based on specific conditions. In theory, AI could also be used in investment decision-making, such as high-frequency trading. That’s not our business, but I imagine others are already doing this.

Ivana Karhanová: What risks do you see?

Vladimír Bezděk: We’re a regulated financial entity. Any AI use must comply with both the Czech National Bank rules and the EU AI Act, which requires explainability and defines risk categories.
The broader question is whether humans will always be able to explain the decisions AI makes. If AI becomes so complex that its creators can no longer understand it, that’s a concern.

Ivana Karhanová: That’s why the EU requires explainable algorithms.

Vladimír Bezděk: And I support that. But the EU is just one part of the world. Other regions may not impose such restrictions, which raises geopolitical and regulatory questions.

Ivana Karhanová: Have you personally seen changes in the job market?

Vladimír Bezděk: Yes. A friend of mine, a certified legal translator, recently told me that his workload had dropped significantly. AI tools now handle much of what he used to do. He’s had to reconsider his entire professional future — and future, and I imagine thousands of others are facing the same situation.

Ivana Karhanová: People once said creative jobs wouldn’t be affected by AI, but even that’s changed.

Vladimír Bezděk: If we look back at history — the industrial revolution, electricity, automation — every wave of innovation has displaced some jobs but created others. I believe AI is the next chapter. At first, it will mostly replace repetitive cognitive tasks, like the ones I described earlier.

Ivana Karhanová: So, you’re optimistic?

Vladimír Bezděk: Yes. But if AI becomes more than a tool — if it starts defining its own rules — then I have big questions. That’s the singularity scenario. If AI no longer needs us to operate, we’re in uncharted territory.

Ivana Karhanová: Should we try to regulate it?

Vladimír Bezděk: Yes, but we can’t stop the development. It’s like nuclear weapons — you can limit use but not halt the science. Every invention has dual-use potential, — for good and for harm.

Ivana Karhanová: Should companies developing AI be held responsible for how it’s used?

Vladimír Bezděk: That’s tricky. We don’t hold knife manufacturers or car companies responsible for how people use their products. I think the real responsibility lies with the user. Regulation should define the boundary between acceptable and unacceptable use —, just like it does with other tools.

Ivana Karhanová: And geopolitically?

Vladimír Bezděk: I think AI development will follow the current trend of deglobalization. Major regions, like— the US, China, and maybe the EU, — will want their own AI systems under their control. We’ve already seen a push for local data centers. Europe will likely face the decision of whether to rely on foreign-developed tools or build its own.

Ivana Karhanová: Will decentralization help keep the balance?

Vladimír Bezděk: Maybe. But if one player gets significantly ahead, — and artificial intelligence becomes more than just a tool, — that could be a game-changer.

Ivana Karhanová: Let’s talk about junior employees. What happens when AI outperforms them?

Vladimír Bezděk: AI is already better at many beginner-level tasks. Those were the very tasks juniors used to learn from. I wonder what they’ll learn from now. I see this with my own kids, who are entering the job market. I believe new roles will emerge, just like they always have — but I don’t yet know what to advise them to study or pursue.

Ivana Karhanová: So, what should young people do?

Vladimír Bezděk: It sounds cliché, but they should follow what truly interests them. Even if it means starting from scratch. Passion, curiosity, and human connection will matter. Routine tasks won’t stand a chance against AI.

Ivana Karhanová: Do you think AI will one day replace you as chairman?

Vladimír Bezděk: Technically, AI could make many of the decisions I do today. But, initiating new things — making bold, strategic moves without full data — that’s still a human role. AI might suggest pros and cons, but the final decision often requires intuition.

Ivana Karhanová: And what about team dynamics?

Vladimír Bezděk: That’s another thing AI can’t do — at least not yet. Human relationships, team chemistry, trust, and — those aren’t algorithmic. In business, success isn’t just about technical decisions; it’s also about how people work together.

Ivana Karhanová: If AI did replace you, what would you do?

Vladimír Bezděk: I’d read more! But seriously, I enjoy learning and sharing knowledge. I’d look for something meaningful where I can still interact with people and keep growing.

Ivana Karhanová: Thank you for the thoughtful conversation.