I’ve always been fascinated by what is possible with data and technology, and data innovation is a thread running through my career. I started programming on a ZX81 aged 9, moving on to running statistics analysis on mainframes in the late 80s, and then working for university research groups as a developer and analyst (what we would now call a data scientist).
Alongside that I was interested in the models and techniques that allow you to use the data and technology in a sophisticated way. My first degree was in theoretical physics, developing simulations to understand galaxy structure, and my MSc and PhD were in artificial intelligence and computational neuroscience, evolving robots to play football (among other things).
A second thread running through my career is applying data innovation to better understand and tackle social issues. Although my first job was hacking data out of government administrative systems to estimate deprivation levels, I found few opportunities after my PhD to apply data and tech innovation to social issues.
So, I co-founded a company to do exactly that, and for over a decade worked with hundreds of public sector organisations across the UK and internationally. Over time I became more involved with government, advising on opening-up and using data, and in 2017 I was appointed to launch the Data Science Campus, growing a team of 70 data scientists and trainers to work across government.
However, that makes it all sound very planned, which it certainly didn’t feel at the time. So, if you’re reading this and thinking about your own career, the main thing I’d say is if you see a good opportunity or a good person or team to work with, then go for it. The only bits I’ve regretted in hindsight are where I didn’t try new things or try to go faster.
Launching and leading a “spin-out” data innovation company from Oxford University, and a decade later handing it over to the next generation of leaders. But the real buzz I get is seeing how the team have smashed it out of the park, while retaining the same open culture and values of innovation, collaboration and social good.
Although leading a group that uses satellite imagery and global shipping GPS data to produce rapid statistics on the environment and economy runs that a close second.
Hans Rosling, who absolutely recognised the power of statistics, data and technology to change the world. Great storyteller, utterly fearless in tackling myths and fake news, and a hugely visible and credible leader.
One thing that happened very much as planned was how we grew the Campus team and impact. Our review of the two years since launch was published in 2019 and looking back it was great to see how many of our original goals we had reached.
One thing that was unexpected was increased understanding of the strength of data science in this country. In 2019, I worked a lot with data science groups across the UK and internationally, and it’s clear that the UK has real expertise and application. One challenge that we highlighted with the Royal Society is how to harness the skills from across industry, universities and government – and that’s something I’d love to work more on.
As a data scientist, it’s hard enough accurately identifying what’s really happening now, let alone trying to predict the future. But there are still a number of things I’m interested in.
The first is growing awareness of data ethics, including the impact of algorithms on fairness and inequality – I’d expect interpretability and output bias to be a huge area of interest and action for industry and users this year.
The second is that the debate on energy use in our work will go mainstream, as we better understand the energy and climate impact of the likes of machine learning model training.
Finally, a wish that the hype around “generalised AI” dies down; leaving the rest of us to get on with the day-job of (responsibly) using data and tech to improve things.
Targeting the biggest social issues facing us, such as climate change, inequality and public health. These are global challenges, and government has a huge role to play in bringing the right resources to bear in tackling them. The moonshot opportunity is to combine data and technology expertise from industry, academia and government, together with the research skills that the UK excels in, and link up with policy, regulation and service design so that we can pull the right levers to make change happen at scale. What could be more important than that?
Legacy. All organisations that have been around for a while face major legacy tech and data challenges, and government is no different. Anyone who can really help organisations crack this is going to earn huge market share; I’d love to see commercial data science/AI groups seriously get to grips with this.
One way of sidestepping this is to enable innovation teams to work with greenfield tech and data estates – the Campus is an approach within government, similarly in market sectors such as fintech you see traditional players supporting start-up/growth incubators. This raises a different problem of how to scale from proof-of-concept to system-wide impact or reversing the innovation back into the wider organisation. That’s the big challenge for us this year.