Why Asking Questions Matters More Than Having Answers

1. AI answers the question you ask — not necessarily the one you should be asking. Knowing which question to ask requires expertise, context, and judgment that accumulates through experience and human interaction, not through better prompting.
2. AI is a probabilistic model, and its output changes with every prompt. Understanding that it is data-driven, math-heavy, and non-deterministic changes how you use it and how much you trust any single response.
3. Expertise cannot be shortcut. You learn to interpret situations, to read what could be right or wrong, to apply a gut feeling built from years of exposure. AI can accelerate information gathering, but it does not build that judgment for you.
4. The colleagues who shape you do something AI cannot replicate. They ask back and reframe your question. They know which answer matters and which problem you are actually trying to solve. That kind of exchange is still where the most useful learning happens.
5. Sometimes the smartest thing is to ask a human instead of a machine. Everyone likes to explain what they do to someone who is genuinely interested. You learn something, they feel valued, and you build a relationship that compounds over time.

When Bettina joined RBI in 2005, Vadym was just a toddler. Back then, some documents were still moved by internal paper mail. And there was a fax machine in the office that was still actively used. Today they work together in the Customer Risk Intelligence team — Bettina as a senior expert with two decades of institutional knowledge, Vadym as a data scientist whose research was recently presented to the Stanford AI professionals' community. That gap in experience, and what each brings because of it, runs through everything their conversation covers.

Bettina's memories are not nostalgia—they are context. Two decades of technology shifts produced an organization that knows how to adapt. "I think a good part is that we really developed the capabilities in the bank to deal with change," she says. However, the pace and nature of change she is living through now are the fastest she has seen in her career.

When Bettina first encountered AI-generated summaries and results, her reaction was immediate. "Oh my god, how can this now be done that easily when I need two hours to come up with something like this?"  Three minutes later, she was generating ideas.

That transition—from shock to application—says a lot. Vadym describes Bettina as someone who has always fearlessly jumped into innovation projects: always the first to think about how a new technology could be useful, ready to experiment before others have decided whether it is worth the effort.

Across twenty years of technology shifts at RBI, that pattern has been a constant. That restless curiosity has left a mark—RBI was recently named Best Corporate and Institutional Bank for AI in Western Europe at the inaugural Global Finance AI in Finance Awards 2025. But ask Bettina how she got there, and she'll tell you it starts with a habit of asking questions and not being afraid to look like an amateur.

Her colleague Vadym, a data scientist, adds the technical frame. The most common misconception about AI, he says, is that people treat it as though it were deterministic, like a calculator: the same input results in the same output every time.

The reality is different. "It is still a probabilistic model that is purely data-driven and has a lot of math incorporated in it that runs continuously just to deliver you some output," he says. The output changes with every prompt, based on how the question is framed.

 "AI might be great at answering our questions, but we still need to be the ones asking the right kind of questions — and we need the ability to assess the answer," they summarize.

Will the jobs students are being trained for today still exist in the future? And what role will AI play in shaping their careers? Bettina emphasizes that while specific tasks and roles may evolve, the qualities that make people valuable are not disappearing—if anything, they are becoming more important. 
 
AI is exceptionally strong at reproducing existing knowledge and recognizing patterns it was trained on. But identifying new patterns, questioning assumptions, and connecting ideas in novel ways still require human judgment and critical thinking. Our ability to think outside the box remains a key differentiator. 
 
As long as people of any age build a solid knowledge base—one that allows them to understand context, spot inconsistencies, and form their own viewpoints—their contributions will remain highly valuable. This is where human input continues to make a real difference. 

The capacity to know what you do not know, to recognize when an answer does not add up—this is built through years of exposure and conversations with people who push back, redirect, and respond to you. "What shaped me the most were my mentors, my colleagues, the discussions I had with them," Bettina says. She points out that AI is built on the knowledge that humans generated together—in dialogues, in training, in accumulated expertise shared across disciplines and over time.

And there is no point in hesitating to “bother” people or being afraid that you’re not going to look smart enough. "Everybody likes to explain their area of expertise," Bettina says. "They feel appreciated when people are interested in what they actually do." The conversation produces knowledge you would not have thought to search for, and relationships that compound over time.

So, practice asking and remember that sometimes it’s smarter to ask a human before you ask the machine.