Women in AI at Oxford · Profile Series

Marina Jirotka

Asking why do people do what they do – and how systems affect the societies that use them.

Read

“We didn’t really understand intelligence, never mind artificial intelligence,” says Professor Marina Jirotka. “And we still don’t.” For her, the novelty of the field connected directly to a question that had long fascinated her: why people do what they do.

Beginnings

Inheriting a scientific curiosity

Marina grew up in Scotland in a family of scientists. Her parents had fled Czechoslovakia before the Second World War, arriving in Britain as refugees.

‘They were all scientists. My mother included.’ Surrounded by a family culture where scientific curiosity was encouraged, her interests extended beyond science alone. What fascinated her most was people.

‘I was always really interested in why people did what they did,’ she says. ‘It seemed to me there were a lot of extraordinary things that people did that I couldn’t understand, so I thought I would try and find out more about why they did what they did.’

She chose to study Social Anthropology and Psychology at Goldsmiths College, a course she believed would provide a foundation for understanding human behaviour both culturally and psychologically.

“We didn’t really understand intelligence, never mind artificial intelligence. And we still don’t.”

The crossover

An entirely new experience in the early days of AI

Marina’s path to computing came later in the 1980s, at a time when there was great excitement in the development of intelligent machines. This was when research focussed on computational models of the mind. Researchers in AI drew on these models to develop Knowledge Based Systems. This was the focus of her first involvement with AI.

For Marina it was an entirely new experience, and the novelty of the field was exactly what drew her in. ‘It was the first time I’d ever touched a computer,’ she recalls. On the first courses she took in Computing and AI there were very few women.

Marina Jirotka in conversation, filmed at the Department of Computer Science, University of Oxford · Watch on YouTube ↗

The research

Knowledge is not just in your head

After completing her course, she joined research projects exploring how computer systems could be designed more effectively for real-world organisations. She came to Oxford to research into Requirements Engineering, where she worked at their Computer Science Lab.

The work drew heavily on ethnography, a research method rooted in anthropology that involves immersing oneself in a particular environment to observe how people behave and interact.

‘Social anthropology really opened my eyes,’ she says. That approach became central to her doctoral research. She spent long periods studying workplaces, using video-based ethnography to understand how employees interacted with systems.

One early study took her into financial trading rooms. ‘That was another eye-opener,’ she says. British Telecom was developing voice recognition technology intended to help traders record information more quickly. By observing how traders actually worked, Marina and her colleagues were able to identify where the technology might succeed, and where it might fail.

“Knowledge is also interactional – it’s not just residing in the head – it is produced with other people in the world around us.”

Future living: an artist’s impression of an AI-powered household robot

The field

Anticipating what could go wrong

As digital technologies expanded, Marina’s research began to focus on emerging online platforms and their social impact, for example studying how children interacted with online environments. This work led to her appointment as a specialist adviser to the House of Lords Select Committee on children and the internet, where she met Baroness Beeban Kidron, a leading campaigner for children’s rights online. Through that role, Marina began to see how large-scale digital systems could draw in vulnerable users and shape their experiences in ways that increased risk rather than reducing it.

This work helped shape work on responsible innovation, which focuses on anticipating both the benefits and the unintended consequences of new technologies.

‘You are trying to anticipate what could possibly go wrong,’ she says. Today Marina applies this approach to emerging areas including AI, quantum technologies, and responsible robotics.

Unlike warehouse robots confined within controlled spaces, social robots interact directly with humans in far less predictable environments. Working with Professor Alan Winfield at the Bristol Robotics Laboratory, the team developed the concept of an ethical black box, similar to the flight recorders used in aircraft. The device records key data about a robot’s behaviour in the moments leading up to an incident.

Importantly, the ethical black box does not determine blame. ‘It’s not the arbiter of truth,’ Marina explains. Instead, it provides evidence for expert human investigators, ensuring that accidents involving autonomous systems can be analysed responsibly.

Also watch Designing, building and using sustainable AI

YouTube

The commercial problem

Gaining a grip on powerful tools

Despite these advances, she remains concerned about the speed and direction of current AI development. From her perspective, one of the biggest challenges is persuading companies and researchers to pause and consider the wider consequences of their work.

The difficulty lies in the commercial incentives driving many technology companies to release systems before fully understanding their consequences. ‘There’s this commercial drive,’ she says, ‘and we have to somehow get a grip on that.’

For Marina, the starting point is straightforward: AI is a tool. What matters is what you do with it.

“We need to understand the implications before letting them loose.”

“AI is a tool. If you take a hammer you can bash someone on the head with it and kill them, or you can bang a nail into building a beautiful house.” – Marina Jirotka, Professor of Human Centred Computing, Department of Computer Science, University of Oxford

29%

of AI conference authors globally are women, a persistent gender gap tracked by the Stanford AI Index

55%

of organisations globally have adopted AI in at least one business function, yet less than half establish clear ethical guardrails for its use

79%

of technology executives agree that AI ethics is a priority, but a significant gap remains between setting principles and implementing them

Behind the camera: filming the Women in AI at Oxford profile series

Women in the field

They want to make a difference

Marina’s career has unfolded within fields historically dominated by men. ‘You had to have a lot of drive and energy and belief in yourself,’ she says.

‘I would say something and nobody would say anything. Then a man would say the same sort of thing and everyone would say, “Oh, that’s a really good point.”’ She believes the situation has improved, and she sees something encouraging in the students now coming through.

‘What I’ve noticed about a lot of the women who apply to study with me,’ she says, ‘is that they want to make a difference.’ Rather than focusing purely on technical development, many are motivated by the potential for technology to benefit society, policy and ethics.

For those considering careers in the field, she is direct about the compromises that can come with it. Finding a workplace where those values actually matter is not always straightforward. ‘Sometimes it takes guts to walk away,’ she says.

Marina Jirotka at the Department of Computer Science, University of Oxford

What’s next

Empowering a generation to make a difference

Today, many of the doctoral students she supervises are women, particularly those interested in technology’s broader societal impact.

For students considering careers in AI or technology research, Marina encourages them to think beyond technical skills alone. ‘Why are you doing it?’ she asks. ‘How does it benefit people and the planet?’

Despite her concerns about the current pace of technological development, Marina remains hopeful about the future. Younger generations, she believes, are increasingly aware of issues such as sustainability, social justice and the environmental impacts of technology.

“The younger generation are much more concerned about how these systems might impact them as they get older.”

“What I have noticed about a lot of the women who apply to study with me is that they want to make a difference and to benefit society” – Marina Jirotka, Professor of Human Centred Computing, Department of Computer Science, University of Oxford