 I'm a political scientist. I look at, I research what politics looks like when it's intertwined with the internet and social media. Because that's what's happening. A growing percentage of people spend growing proportions of their time online. That's long been true of the developed world. It's increasingly true of developing countries as well. So what does political participation look like on the internet and social media? Basically, tiny acts of political participation become viable. People entertain themselves online, educate themselves online. They shop, work, bank, date, borrow, steal if that's what they're into. And they come across all sorts of opportunities to participate politically. Clicking like, changing a status, joining an email campaign, signing an electronic petition, tweeting, retweeting, etc. You know them all. You're exposed to them all in your daily lives. And people seem to be taking up these opportunities, even groups that we traditionally think of as not participating. Young people, for example. These sound insignificantly small. They sound like nothing. But these tiny acts scale up to large-scale mobilisation. We've seen it in authoritarian states in the revolutions of the Arab Spring, for example, where key websites gradually built up resistance against the regime that tipped over to key demonstrations like this. We see it in democratic states where mass email campaigns and electronic petitions have gained millions of supporters which have really brought about policy change. But because we hear so much about the success of social media and mobilisation, we forget that, actually, most mobilisations fail and we never hear about them at all. One robust finding across countries is that 99% of electronic petitions go absolutely nowhere. And we don't know much about why the ones that succeed do succeed. In the UK, for example, petitions started at exactly the same time about the same issue. How to stop the culling of cute badges like these have had completely different fates. One of them bombed out completely. One of them got huge numbers of supporters. There's other things we don't know. Because mobilisations like this can be going without the traditional trappings of mobilisation, without leaders and institutions, for example, when the Brazilian president asked to speak to the lead in that demonstration, she was told, there aren't any leaders. They're characterised by tipping points, gradually rising and then tipping over into critical mass. So mobilisations like this are unstable, they're unpredictable, they're difficult to understand. At the Oxford Internet Institute we've developed a model of chaotic pluralism to try and encapsulate this. It's a real challenge for states to understand and predict. But the good news is that social media also provide the solution. Every tiny act leads to a digital trace like this. Those digital traces can be mined to generate what we call big data. Big data, real-time transactional data of a kind social science has never had before. It's the kind of data that natural scientists have, people like Chris. We can start to use that to understand social systems in the same way a scientist have understood natural systems. Here a member of our research team, for example, has used Wikipedia page views to predict the last Iranian election more successfully than any conventional method of analysis. Some people have argued that just as meteorologists have got better at predicting the weather, political scientists can get better at predicting politics. Here, because meteorologists were able to predict the weather and the floods that overtook that Oxford two weeks ago, quite close to where I live, the city of Oxford was able to react more quickly and provide better flood defences than it usually does. Understanding chaotic systems like this can lead us, can lead policymakers to be able to make better environmental policy. The read across for governments and political institutions is that by understanding political mobilisation, like these snowmen against globalisation, and being able to understand their concerns, their experiences, their complaints, their preferences, their behaviour and their needs, they may be able to design policies that are actually a far better reflection of what's going on in their country than previously, be able to kind of see like a citizen rather than seeing like a state.