Henry Baker didn’t take the typical path into data science. With a history degree from Oxford and six years in Japan working on technical and educational issues related to Sub-Saharan Africa, his story is a testament of curiosity, adaptability and the power of interdisciplinary thinking.

“I’m just interested in how the world works,” Baker, 32, told me in Nagasaki, explaining his shift from studying modern history to researching data science. Now a data science researcher at the Global Public Policy Institute (GPPi) in Berlin, he sees the two fields – though seemingly worlds apart – as parallel approaches to the same question: why is the world structured the way it is?

“History tries to interpret meaning through qualitative analysis, while data science quantifies that same complexity through numbers.”

A surprising connection between the two came to him via science fiction. As a teenager applying to college, he wrote his Oxford personal statement on Isaac Asimov’s Foundation series, particularly fascinated by the concept of “psychohistory” – a fictional discipline combining statistics and sociology to predict the future. Years later, he realised that psychohistory foreshadowed the core aims of modern data science: modeling the world through agent-based simulations.

Before immersing himself in data, however, Baker took a detour – one that led him across the planet to Japan. He moved to Japan for a job focused on East African education access, working with NGOs in Uganda and Senegal. “I didn’t come to Japan for Japan,” he said, “but I ended up staying almost six years.”

Living and working in Japan shaped his views profoundly. “It was my first major job, and I internalised a lot of the Japanese work ethic, sometimes to my own detriment,” he said, citing long hours, rigid hierarchies and a culture of presenteeism.

Baker emphasised that these experiences, though seemingly unrelated, all contribute to his current research. “Oxford taught me how to learn independently. Japan taught me resilience. Data science lets me ask new questions about the world.”

He also credited his visual thinking style with helping him grasp complex mathematical concepts, particularly linear algebra, the foundation of many machine learning techniques.

“Follow what genuinely interests you”

On the topic of AI, Baker offered a unique view. While many experts predict exponential progress, he suspects we may be reaching a plateau with current large language models.

“There’s only so much intelligence you can extract from language prediction. It might sound human, but that doesn’t mean it thinks like a human.”

Unlike Asimov’s vision of Psychohistory, today’s AI is not yet capable of using data science to predict human society.