John Taylor, new President of the IFoA, shares his views on the ways in which the actuarial world is being affected by the inexorable rise of data science.
It’s been suggested that, in their approach to Big Data, actuaries are ‘the original data scientists’. Could you explain that connection? What, in their essentials, do actuaries have in common with data scientists?
When I heard actuaries described as the original data scientists, I really liked the phrase. It captures the essence of what actuaries have been doing for centuries. If you look back at the pioneers of the profession, they were analysing what at the time were large data sets to do with longevity, trying to find the truth and insight in them, and then extrapolate into the future. Then if you look back to just a decade or two, actuaries did similar things around Stochastic asset models – and, again, trying to make projections into the future. And that’s the essence of what data scientists try to do. The tools and technology used have moved on massively, of course, but the fundamentals of trying to extract truth and insight from data, and then using that to infer something about the future, are the same.
Why are actuaries so well placed to explain data science? And can data scientists learn from actuarial practice?
What actuaries bring is a sense of business context, commercial value, and of what it all means for their client – be it a pension scheme, an insurance company, or indeed the ultimate end-user, the consumer. Having that sense of business context, as an overlay to the raw insights that a data scientist may bring, makes those insights more useable and valuable to the client.
The IFoA plans to launch a data science certification. What will be its scope and what will members who undertake it gain?
There’s an appetite among IFoA members to learn more about data science, and so we are about to launch a certificate in data science, available to all members of the actuarial profession. They’ll be able to choose from a set of five modules that cover disciplines such as data visualisation, Artificial Intelligence and Machine Learning. They can pick and choose from them, and those who complete the set of five will be awarded the certificate. From this, actuaries will get a much greater appreciation of the art of the possible in data science. It may not make them data science practitioners per se, but it will enable those certified to work more closely with data scientists. And the certificate will also have a currency with employers in evidencing the additional learning that actuaries have undertaken.
The traditional fields for actuaries have been insurance and pensions. Now actuaries are appearing in agriculture, healthcare, climate change initiatives, and even ride-sharing app development. How does this reflect the way data science is developing?
Data science is very empowering for actuaries. Not only does it bestow the ability to re-engineer what they do in traditional fields, but it gives them a platform to move into wider fields. The underlying driver behind this is that other industries, that haven’t been so accustomed to dealing with data, are now being deluged with it. And they don’t really quite appreciate how to derive value from it. Bringing-in a profession that’s long been expert in dealing with that data, and getting insight out of it, is a tremendous opportunity for those industries – even some of the creative industries like advertising. The ‘Mad Men’ are being replaced by the ‘Math Men’, as advertising becomes much more data-driven about defining individual segments and offering bespoke propositions – and actuaries are getting involved in that kind of innovation.
You have spoken about the ethical implications of data science. What are the key considerations around this question, where actuaries can bring thought leadership?
Developments in data science are throwing up some challenging ethical issues. That does create an opportunity for a profession with an established code of ethics to step in and try and make a difference. The IFoA is also working with the Royal Statistical Society to bring forward a code of ethics around data science. And we’re working with government and policy-makers to make sure that the evolving policy keeps up-to-date with data science.
‘Data science’ is often referred to in public debate, but definitions of what ‘data science’ actually means seem to change continually. How well understood do you feel the term is?
I’m not sure that ‘data science’ is a terribly well understood phrase, in fact. I was speaking to an actuary about my ambitions for the IFoA to do more in the field of data science, and his retort was ‘That’s great – but we also need to be doing more on Artificial Intelligence’. Well, actually, I was using a definition of data science that encompasses AI, so that demonstrates that there’s a degree of confusion around the term. It’s such a dynamic field that as soon as you try and pin down what it is today, it’s changed tomorrow. So, I’m fairly relaxed that it’s dynamic and evolving term.
What do you believe makes the contribution of the actuarial profession so uniquely valuable to the emergent field of data science?
As professionals immersed in business context, and in understanding the impact companies can have with end-users, actuaries are well-placed to add value and utilises the insights data scientists can generate. Let me give you a couple of examples. I know an actuary who works in insurance, and who is looking at new ways to deliver service to existing clients, and that’s about understanding individual customers at a very small level of detail, and looking for trigger points when they might be receptive to new service offerings. This approach is really transforming the take-up rate for services. Second, I heard a compelling example at a conference, where actuaries working in the field of marketing were taking huge amounts of data from social media, and again, trying to infer which customer segments might be suitable for which offerings – and at what times – and then making them available in almost a test-and-learn environment.
How important will collaboration between actuaries and data scientists be, going forward?
Collaboration is going to be vital both for the IFoA and for individual actuaries. Now, most actuaries will be very familiar to working in multidisciplinary teams, they’ve done that for a long time, but I believe they will increasingly do so alongside data scientists and other technologists. The blend of skills will make those teams very powerful. But it’s important that they are able to speak the same language as one another.
What would you say to actuaries who have forebodings that some data science may prove a future threat to their expertise and livelihoods?
It’s very hard to say what the future will look like in, you know, a hundred years’ time. What I can say is that, for the foreseeable future, I think that there’s a fantastic opportunity for actuaries because of data science. The fact that data is proliferating in not just traditional industry sectors, but everywhere else, means that there’s a kind of crying-out for a profession that can understand the value of data and make predictions based on that understanding. That’s where the essence of the opportunity is for us, I believe.
Events calendar
No results found.