INTERVIEW – Kateřina Lesch: Social Stereotypes Still Influence Women's Career Choices

Kateřina Lesch was one of the speakers at the autumn conference Women in Technology, organized as part of our platform for women Future is Female: link here

Technology and artificial intelligence are among the fastest-growing fields today, yet gender diversity remains a persistent challenge. How can women contribute to AI development? How can we encourage more women to enter technical fields? And what is the role of so-called soft skills in working with AI?

We discussed these and other topics with Kateřina Lesch, data analyst at EmbedIT, mentor at Femme Palette, and advocate for diversity in the tech sector. Kateřina shares her practical experience, reflects on challenges related to AI diversity, and reveals which changes in the field could ease the inclusion of people from diverse backgrounds.

“Due to social stereotypes, it can sometimes be difficult for girls to choose, for example, mathematics or physics. That’s why I try to lead by example — to show that you don’t become an introverted bearded guy the moment you enter the faculty, and that you can gain useful and career-relevant knowledge,” says Kateřina Lesch in the interview.

INTERVIEW

How are you working at EmbedIT or through other projects to encourage women to participate in data science and AI?

At EmbedIT, we mostly benefit from already-trained female data scientists and tech experts. But their preparation starts much earlier. Due to social stereotypes, it can sometimes be hard for girls to choose math or physics. I try to set an example — to show that entering the faculty doesn’t mean you’ll become an introverted bearded coder and that you can gain valuable, career-oriented knowledge. I often talk about this at conferences and mentor women interested in data science through Femme Palette. EmbedIT helped by giving me the opportunity to work in a responsible and interesting data role.

Do you see specific examples in AI applications today where gender diversity leads to better results? If so, what are they?

Unless gender diversity is present in the training data, we can’t expect it in the applications. On the other hand, training data reflects societal reality, and I think artificially manipulating it in favor of any group could be dangerous — bordering on social engineering. We would have to ask ourselves how much to inject, in favor of whom, and who makes that decision.

Do you have any specific examples where team diversity directly influenced the success of an AI project? How do you assess such situations?

Not really. AI is impartial and doesn’t care who trains it. From my experience, both men and women successfully define AI use cases — sometimes in balanced teams, other times not.

Do you work with any specific organizations that support women in IT?

I’m a mentor in the Femme Palette program in the tech field. I have three children, so I’m also interested in combining an IT career with motherhood. I used to work with Czechitas Digital Academy, though I no longer have enough time for it.

How important are “soft skills” in AI, and which ones are most in demand, especially for women looking to enter this field?

Key soft skills include open communication, bringing together representatives from various teams and domains, helping define meaningful AI use cases, and identifying business cases. Creative language use and empathy are also valuable when prompting.

What do you see as the biggest challenges related to diversity in AI and tech, especially compared to other technical fields?

According to research, AI is replacing junior positions in tech — which are, unfortunately, often held by women. The challenge is finding meaningful roles for them.

What specific measures or approaches could the European Commission take to fulfill the “Gender Equality Strategy” and improve the status of women in AI?

I’m not entirely sure, but we need to start teaching tech equally to both genders in elementary schools. Every child should have the chance to choose their path without feeling that a field is “not for them” because of gender. I think we’re getting better as a society. Still, the AI curriculum changes rapidly with the field’s development.

How should AI be regulated to minimize bias and stereotypes that may negatively affect women and other groups (if you think it should be regulated at all)?

There are definitely areas where AI should be regulated to prevent misuse. But regarding gender bias, I’m not so sure — AI systems reflect their training data, as I mentioned above.

How do you think the qualifications and skills required in AI will evolve to make the field more inclusive for people with different educational and professional backgrounds?

I think that’s already happening — the current boom favors newcomers and people from non-tech backgrounds. AI is accessible to anyone; all you need is curiosity and a willingness to play with new tools. I know many people, especially women, who transitioned to AI power users from completely different fields, like marketing. In the future, the key qualifications will be creativity and a willingness to explore new things.