Data and Bias

Data has an unearned aura of objectivity. Statistics certainly has the power to illuminate, but when collected without care, it can also exclude and discriminate. Our insatiable appetite to turn the world into something that can be counted can force a gap between what matters and what can be measured.

In this talk, with wit and warmth, I want to look at some of the ways that bias has become such a profound modern issue, how it can be amplified, the ways in which it causes genuine harm and what we can do about it. I’ll look carefully at the issue of fairness – how it can be defined and whether it can be achieved. And I’ll build a persuasive picture of why who can and cannot be counted is one of the defining issues of our days.

What data can (and can’t) tell us about ourselves

In the era of Big Data, we’ve come to believe that, with enough information, human behaviour is predictable. And to a large extent, it is. There are no shortage of terrific stories, where numbers have unlocked the answers to our biggest questions.

But numbers can also lead us perilously awry. Science is no stranger to this fact. Some of the best in the world have been guilty of finding signals in the noise where there are none. Of missing and misinterpreting subtle signals that go on to have dramatic and catastrophic consequences, of falling into the trap of over-relying on the numbers and of believing them to hold an objective fact.

In this talk, I want to take you on a tour of some of the most important lessons science has learned in 2020 and beyond. I want to show you why they matter, to share with you some extraordinary stories about what happens when things go wrong and to probe at the edges of quantitative thinking. We’ll decide what parts of our future are truly forecastable, demonstrate the awesome power and potential of data, and find its limitations.

The trouble with automation

We like to think of ourselves as master decision makers; as perfectly rational creatures, grounded in reason and logic. It’s a nice idea, but the reality is rather different. In truth, humans are a mess of competing incentives, of bad memories and of blind, impulsive biases.

The modern era of data analytics is, in some sense, an attempt to automate our decision making, iron out some of the fallibilities and biases built in to our choices. But decisions driven by data have blind spots too. And all this leaves us to a conundrum: humans are flawed, machines are flawed. So who do we want to leave in charge of our decision making. Who should be the ultimate arbiter when there are very real dangers of leaving any one side in charge?

This can be a session with lots of audience interaction, that aims to uncover some of the hidden flaws in our own decision making, as well as those of the systems we’ve built to replace us. It aims to explore some of the greatest ironies of automation, and puts into perspective some of the hardest challenges facing the future.

The Joy of Data

I’ve spent the last decade working with data, hunting for mathematical patterns in human behaviour. In that time, I’ve come across some incredible stories - written solely in the numbers - that get right to the heart of who we are as people.

In this optimistic talk, I’ll share some extraordinary tales about what’s happening at the very cutting edge of data science – a host of surprising and delightful stories that demonstrate how far you can go when we look back at ourselves through the eyes of data. And I’ll show you how a mathematical view of what means to be human can shape the way we design our society, from dating and healthcare to catching serial killers and everything in between.

The winner’s handbook

Only an idiot would choose to play chicken. No sane person, surely, has ever chosen to drive directly towards their opponent at speed, knowing that choosing to swerve away from an inevitable collision will result in a loss.

An idiotic game, yes. But not one without a winning strategy . Indeed, ask a mathematician, and they’ll tell you the best way to force your rival to blink, is to unscrew your own steering wheel, wind down your window and throw it into the road.

Welcome to the world of game theory. The mathematics of winning. In this interactive, high energy talk, I want to give you a flavour of what game theory has to offer – the surprising and counter intuitive strategies it sometimes suggests, and the difficult questions it asks. To be a good negotiator, do you have to accept some risk of disaster? And – perhaps more importantly - is there always a way to get what you want?