Dr Anne-Marie Imafidon MBE delivered a Turing Lecture on AI and the future of work on 26 February. She recently sat down with Jessie Wand and Beth Wood from the Turing to give her personal perspective on three big questions. The following interview has been edited for clarity.
Anne-Marie was the youngest girl ever to pass A-level Computing at age 11, and was just 20 years old when she received her Master’s Degree in Mathematics and Computer Science from the University of Oxford. Since then, she has held positions at Goldman Sachs, Hewlett-Packard and Deutsche Bank and obtained Honorary Doctorates from Open University, Glasgow Caledonian University, Kent University & Bristol University and an Honorary Fellowship at Keble College, Oxford. Her wealth of experience and pioneering spirit led her to co-found the STEMettes, an award-winning social initiative dedicated to inspiring and promoting the next generation of young women in the STEM sectors. Since its inception seven years ago, it has given more than 45,000 girls across Europe access to Anne-Marie’s vision for a more diverse and balanced science and tech community.
What to you are the most exciting innovations in the fields of data science and AI that you foresee coming to fruition in the next decade?
Anne-Marie Imafidon: I’m most excited by the tech for good movement and the notion of responsible tech becoming more and more mainstream. The optimist in me is excited to see how—post-Brexit—the UK could maybe become, if not a leader, then one of the pioneers in tech for good. This would involve skilling up and mashing our culture with technology itself, and I think we’re quite uniquely placed to do this.
Currently, responsible tech is in its early stages so I’m quite excited to see what happens when we’re able to use the tech to genuinely solve problems. For example, Doteveryone started using technology to help with end of life care. I think there’s so much more scope for us to solve real societal problems with the data sets we’ve got, but as it’s not necessarily commercially viable, it’s not been prioritised or been done.
You just need someone with the skills who either has spare time or isn’t solely driven by money who is able to say, “I’m going to use this for mental health or to close the race pregnancy gap in the NHS.” I think there are so many clever things we can do with data that I’m genuinely excited to see how the next 10 years go, and the more people we train for this skillset, the more we’re going to see it happen.
What single thing do you think we can change to keep more women working in STEM fields?
If I had to suggest one thing for retention, it would be to enforce compulsory shared or equal care leave, which includes parental leave, across the tech industry if not wider. That’s because not all women have children. Some do and some don’t and it’s completely their choice and their partner’s choice, but one of the biggest things we have impacting attrition and retention is this sense that most of the care that’s done at home is the responsibility of women.
So, there’s an assumption in terms of decisions around promotion and recruitment and even responsibility levels. For instance, big ticket projects that are given or not given to women depending on whether you think they’re about to have a baby or whether they’ve just come back from having a baby or whether you’re worried they might get pregnant at any point.
I think what ends up happening—one of the big things again, if I can choose only one—is that women end up not being valued or promoted and so leave. It’s almost as if the system allows that, if it’s jokes being made at their expense or harassment etc, the onus falls on women.
By opening up shared leave, you allow people to consider a different perspective and work more flexibly. I think this is where we’re going anyway with the future of work but at least it might accelerate us towards equality and how we view this as employers, as managers and as organisations. This change will also help managers deal with a new norm, where it's not only women taking time out to care.
Are there any jobs currently being done exclusively by humans that you think a robot would do better?
Are there any jobs left that are done exclusively by humans (laughs)? It’s a tough one because you have to be careful with what you define “better” as. Is it faster? More accurately? What does better genuinely mean, not just for your bottom line, but also for society?
I think there probably are lots of things that robots could do better if we can program them but we're quite far off doing that in a lot of places. There are parts of almost everyone's job that could be automated or where they could maybe even have their own robots that do elements for them or alongside them, but whether that entire job should disappear is something I'd question. Because ultimately you're going to end up being replaced by a robot that doesn’t work 100% of the time because technology often sucks and then you’re going to have to go back to humans anyway.
However I think the legal profession, for example, is a strong contender where there’s much that's done manually that doesn’t need to be by humans. If we look at scanning documents, analysing and trawling through them, I think there's quite a lot that robots could probably do. There are lawyers where it might make up the bulk of what they do but it’s not everything.
There’s a great DoNotPay algorithm for appealing parking tickets, for example. I didn't know that parking ticket lawyer was a role that a human did! I think anything that is kind of formulaic there’s scope for robots to possibly do this better. But only a few things in life are formulaic…life’s imperfect, it’s not maths.