“What gender is AI, Nadine Bienefeld?”

The doctor of psychology and Federal Institutes of Technology (ETH) lecturer in human-AI interaction talks about prejudices and opportunities: part two of the interview series on role models that promote equality.

It’s not just about structures, but also about behaviour. Many women doubt themselves. Every now and then, an air of confidence is required, especially when it comes to technology.
Nadine Bienefeld

What would you say if I told you that I had these interview questions written by ChatGPT?

Well-structured working-time efficiency. However, I would expect that you have reflected critically on this.

How sexist is AI? It is considered objective – but does it actually reproduce patriarchal structures?

By definition, it is not objective. AI is based on data – it reflects this, including social biases. Discrimination, including sexism, can become entrenched in the system. The example of application algorithms has shown that, based on historical data, patriarchal structures have been reproduced and white males preferred. However, I see an opportunity: if we work transparently, we can uncover biases and consciously counteract them through fine-tuning – with all forms of inequality, not just gender, which is particularly important in the case of medical AI. AI is neither good nor evil, but it is malleable.

So do we need feminist AI?

No. That itself would be exclusive. We need fair AI for everyone.

Is it enough to get more women into AI development?

It’s not just about structures, but also about behaviour. Many women doubt themselves. Every now and then, an air of confidence is required, especially when it comes to technology.

Women often have inferiority complexes in maths and technology...

It depends on experience. My male maths teacher said at the time: “Will the blonde girl next door understand it?” Top marks in statistics or the title of professor can only partially heal this dent to self-esteem.

You say that AI is a great opportunity, especially for women.

AI can automate analytics and software development. This makes human skills such as empathy more important, because they cannot be automated. For example, care work, which is traditionally associated with women, can gain greater visibility and recognition.

What gender is AI?

We have the dangerous tendency to humanize it. AI is neither male nor female – it’s simply technology.

Where does AI assist you in your everyday life?

Outside of work, not at all – it has no place in relationship matters, where many use it. At work, it has become indispensable.