An interview with Mathematician and UCL Lecturer Hannah Fry
Love and math. You almost can’t think of two worlds further apart. However as Hannah Fry proved at the Analytics Experience in Amsterdam, mathematics can uncover the secret of a successful relationship and predict divorce. We had the opportunity to talk to the lecturer in the Mathematics of Cities at the Centre for Advanced Spatial Analysis at UCL, and asked about the relation humans have with new technologies like AI and how it can impact society.
The main message Fry had for the 1400 data and analytics professionals that gathered in Amsterdam was: if you let the data lead you to the insights, there is no telling what you might discover. “It’s easier than ever to uncover patterns and to dig into the why of human behaviour”, Fry states. “New technologies like machine learning have made predictions much more accessible and possible. In fact I see people trying to predict human behaviour in every context. For instance in the US they’re using predictive analysis to determine if a judge should set bail for a suspect. You can argue whether that’s a good thing.
Humans versus algorithms
Fry believes that even though technology provides us with endless opportunities, some things should be left for humans to decide and judge: “The discussion we should have in society is to what extend we let algorithms determine what we should decide or do? Do we want to leave our future to algorithms? This is a really important question as we will never be able to predict everything a 100 percent right.”
It’s also a question that can’t be answered with a simple yes or no. It should be judged per case and based on to what extent it impacts human beings. “Take this example in China where they use facial recognition to stop people from taking too much toilet paper. The dispenser gives out 60 inches of paper for each person. If it notices the same person is in there within the next 9 minutes, it locks the dispenser. Not a big problem if the system fails once in a while. However in healthcare the wrong decision can be a matter of life and death. Can the algorithm predict with a 100% certainty that this tumor will grow into cancer? And what are you going to do when the algorithm was wrong and you’ve suggested the wrong treatment to the patient? These are issues that need careful consideration.”
AI in dating 
As Hannah Fry is also the author of The Mathematics of Love, we’re curious to hear if and how AI will affect dating. “People have been trying to work out why people like each other for ages. And there really is no real data or technique to predict this. But it gets a bit easier if people actually see each other and you can analyse how they respond to one another. A data scientist in the US has applied machine learning in speed dating sessions through Skype. Based on analysis of the facial expressions he predicted the compatibility. This is an interesting start, but I think also online dating will become more and more clever.”
Will AI take over my job?
AI doesn’t only generate positive vibes, it also emcompasses fear. People are worried that smart, applied AI will take their jobs. Fry: “Sure there will be some professions that don’t exist anymore in 10 years from now. But people are underestimating the value of the human touch. Social workers can never be fully replaced and in depth journalism is also a good example. The output of AI is based on information and data that’s already there so it will not generate a truly original background story about Brexit or the elections. In fact I’ve been to a musical that was written with AI technology. It felt like I’ve already seen it and knew the story. That’s because it was based on success factors of existing musicals.”
Fry is convinced that people tend to overestimate the influence of tech in short term and underestimate it in the long term. “Driverless cars are the perfect example. These would only work in a closed environment otherwise they would still be confronted with cars driven by humans. And as we all know human drivers are unpredictable. The technology can’t deal with this yet. Or you would have to make sure everybody has a driverless car at once. There is no transition period possible so before we can truly leverage this innovation on a big scale, people need to think about a solution for that first.”
Hello world
If you’re as intrigued by Hannah Fry as we are then stay tuned for her new book. ‘Hello world’ will be published in 2018. The book describes how algorithms are infiltrating every aspect of our life. Furthermore she talks about the question where to draw the line between human messiness and the comfortable predictability of a data driven world.
