Predicting
human
behaviour
with AI

Growing up, did you always think you would do something entrepreneurial? In my world, there was no notion that you would found a tech start-up but it was very normal to start something new. For example, my parents were originally selling clothes on the street in Chengdu, the capital of Sichuan.

That kind of enterprising thinking is the ambient temperature in China. We don't call it entrepreneurship, we just call it 'life'.

Who had the greatest influence on you growing up? My mother was a big influence. She had an insatiable desire to learn and would take me along to classes when I was very young. She is a descendent of Mencius, a disciple of Confucius and we always had it impressed upon us that we needed to live up to the ancestral name.

But I didn’t spend all my time immersed in Chinese tradition. As a child, I was also a sci-fi nerd and read massive amounts of Western science fiction, along with biographies of famous Western scientists and inventors, such as Benjamin Franklin, Ada Lovelace and Steve Jobs.

It was through my mother and her appetite for knowledge that we met my two Shifu, or teachers. I play the Guqin, a Chinese classical instrument, and my musical Shifu showed me that there is a world of beauty and aesthetics that transcends the mundane.

My Tai-Chi Shifu looked after a different side. He was an adventurer who encouraged me to take risks, to go do things, and I think I got a bit of that energy from him. Together, they gave me a very good balance.

When you were 14 you left home and moved - by yourself - to a school in New Zealand. How did that come about? I always had this instinct that I wanted something more from my education. Maybe it was the influence of my two Shifu, or maybe it was because of my sci-fi obsession, but I wanted to be in an environment that would give me more than just academic success.

My parents were understandably resistant to the idea. It is hard to let your son go at that age and there was a lot of uncertainty about whether I would be led astray or become stronger because of it.

Moving to New Zealand was clearly a very formative experience for you. What are the main things you took away from your time there? In China, you have one path — to study well. When you are 15 or 16, you choose to become either a good technical student or a good arts and humanities student. That choice essentially determines the rest of your life. It’s like being on a train.

When I went to New Zealand, suddenly I wasn’t on that train any more. Or in a race. Everyone was very laid back, and they didn’t seem to have any expectations of you.

Without expectations, everything is possible. But in that realm of infinite possibility, I began to wonder what I was going to do. And that’s when I realised that I could be the author of my own destiny.

People didn’t really expect me to study literature. If you are an international student, you are expected to do English as a second language, but I wanted to do Shakespeare. I wanted to learn poetry. It was hard, but I loved it.

I was an intense child and arguably a workaholic. My friend’s father said to me: "You are running very fast but sometimes you need to pause and feel the sunshine and see the flowers". That piece of advice stayed with me, but it was a long time before I acted on it.

How did you end up in Cambridge? Cambridge was the moonshot. It was an audacious defiance of other people’s expectations. The path I was on was to go to university in Auckland, do an engineering degree, get a job, get the visas sorted and build a life there. That would be the dream of any immigrant being able to settle in a place.

But I had the grades, and I began to ask myself: why am I stopping myself from applying to these universities? I went ahead with an open application to Cambridge and Selwyn offered me a place to read Education.

Why did you want to study education? Alongside my schooling, I was running an education business in New Zealand and I spent a lot of time thinking about the impact of different types of education on people’s lives, including my own.

I had been given a chance not to stay on the path set out for me and I wanted other people to have the same opportunity.

I was thinking and writing about these questions and had built up a following in China of people who were interested in my work. But when I arrived in Cambridge, I soon realised the course was more abstract than I had anticipated. I was learning about the theory of knowledge, while I had thousands of parents in China asking me how to lead a better life.

It was this gap that led me to switch to behavioural science to better understand how individuals influence each other and how individual psychologies relate to cultural psychologies.

Because I was a physics and maths student back in high school, I did a lot of coding. Naturally, that drew me to thinking about how we use computers to model the process of individuals influencing each other on social media.

That sounds like the beginnings of Artificial Societies? In fact, the idea for that started during COVID. I was bored, wandering around the computer science section of the library. A book jumped out at me on social networks, crowds and the sharing of information and that set me on my path. It’s funny how unpredictable life is, particularly when it’s now my job to predict it.

How did being at Cambridge help? The most powerful thing about Cambridge was the access to people. The fact that I could just email academics who were at the leading edge of research in these spaces meant that, in my second year, I was second author on a paper that defined the frontier of computational social science.

I’m still very much in touch with two of those academic mentors. They are advisors to Artificial Societies, and we still have an academic collaboration today.

Can you explain what Artificial Societies does? We predict how a group of people might react to information. For example, we help large companies anticipate how people they care about - such as their customers, shareholders, or opinion leaders - might react to different decisions that they make.

What is the most important thing you have learned about being a CEO? In the beginning, you try to do everything yourself, but you eventually realise that there are some tasks that the team is best placed to take on. I have a highly capable team: I needed to let go and trust them.

You also have to understand that you all care deeply about the business, but people manifest that in different ways. That was a very important lesson, and it is, I think, one of our strengths as a team.

You communicate through listening, by asking good questions and by trusting your team.

One of the things that makes Artificial Societies strong is just how high-trust the environment is. We can count on each other, and, if we disagree, we will listen rather than fight.

What will success look like for you? We want to build a technology that is useful for the world. You need some sort of feedback mechanism to tell you if it’s useful. Revenue is one of those mechanisms, but it’s a brutal metric.

If you build something and you can’t find someone who wants to pay for it, the chances are it’s not useful to them; it’s not solving an important enough problem.

That’s one aspect of it. The other is that this technology has the potential to allow us to understand in advance how whole societies might react to changes.

I’m really excited about the possibility that we could predict how things will play out before making a major decision, such as changing interest rates.

What if there is a way to make Britain even more innovative than it is now? Today, we only have one shot at these things. We can try something and see if it works or fails. But if we can experiment first in what we are calling a ‘societal world model’, maybe we can increase the chances of getting things right first time.