Discover more from The Ruffian
And she'll always get the best of me, the worst is yet to come
This week: what's in a face, how fake news really works, and Putin's surprise.
THE FACE PROBLEM
In recent years, a group of businesses have started providing something called "emotive analytics. It's a way of using technology to detect and classify emotions in consumers, at scale. It has designs on marketing, healthcare, education. Emotions are obviously big drivers of behaviour, so it's useful to know what your consumers or users are feeling, and when. The data comes from our faces, via photos and videos online, from our words in social media, and from our bodies via our Fitbits and Apple Watches. Facial recognition is a big part of it - here's a company that focuses on just that. The idea is that algorithms can identify certain expressions and connect them to the emotions they signal - joy, fear, anger. This is all rests on research from psychologists in the twentieth century who claimed to have established that there are universal, trans-cultural facial expressions of emotion.
A psychologist called Lisa Feldman Barrett has bad news for these companies: their whole model is based on flawed science. In her fabulous TED talk (and this hour-long lecture) she explains that there are no universal facial expressions. The old scientific consensus has been comprehensively destroyed by new evidence. It's now clear - as perhaps it always ought to have been - that the same facial expression can mean several different things, and that the same emotion can be expressed in different ways. A picture of a face that looks fearful in close-up is revealed to be one of pleasure when you see she's drinking a chocolate milk shake. This isn't just about faces, either: the truth is you can't reliably interpret the emotional significance of what a person is doing or saying without understanding the context in which they're doing or saying it. Technologists are always trying to strip out context, because doing so makes data-gathering easier. But just because something is hard to measure doesn't mean it's not important. (We all have a habit of over-confidence in our ability to read faces - something I wrote about here, with regard to the Amanda Knox case; now these companies are reproducing that mistake at scale). Anyway - take a look at the talk, it's about a lot more than faces. It offers a new way of thinking about emotion itself.
THE FAKE WE
Dan Kahan is one of the leading experts on the psychology of political polarisation. Here's a summary of a paper he recently presented on fake news. It makes an interesting argument. When we talk about fake news we usually mean the dissemination of made-up news stories: lies. For example, before the US 2016 election there was a story that Pope Francis had endorsed Trump. But this is only one way fake news works, and Kahan thinks it's the less important one. The other way is the dissemination of fake social proof. Social proof describes how we feel more confident about our views or behaviours when we know that others share them. The Russian propaganda operation is skilled at creating memes that are vehicles for the spread of social proof on extreme positions ("Like and share if you want burqa banned in America"). This is a far more effective technique, because it works with the grain of its audience. Instead of trying to deceive, it taps into people's tribal loyalties. It makes people feel better about taking extreme or angry positions, because they can see others doing it. Russia's aim is not to create false facts so much as to poison political debate.
WORLD VS PUTIN
Somewhat under-remarked, maybe because it's hard to believe: Theresa May and Boris Johnson have done a very effective job on something. That is, putting together an international coalition versus the Russians, over Salisbury. As the excellent Shashank Joshi explains here, it's partly because of timing. This happened just at the point that much of the world was getting fed up with Putin's misbehaviour. So the reaction to it has been bigger than he probably anticipated. Remember, Putin is not a strategic mastermind, and Russia is not in good shape (Obama's unnecessarily brutal assessment was essentially correct). His smart-ass tactics in Syria have created a quagmire for Russia's military. The question now, as Joshi says, is whether Putin's renewed isolation over Salisbury will make him draw in his horns, or lash out.
I did not know about this sublime version of Blackbird, by Crosby, Stills & Nash (or this mind-boggling one by Bobby McFerrin) until today. They're mentioned in an episode of Screw It We're Just Going to Talk About the Beatles, which just happens to be my favourite podcast in the world. I keep meaning to write about it but I'm scared I'll write 5000 words about why I love it so. Let me get back to you when I've got it down to a couple of thousand.
As this tweeter remarks, it must be the most stunning police mugshot ever (oh and there's a less beautiful sight here).
If you're enjoying The Ruffian, please generate some social proof by spreading this link: https://tinyletter.com/IanLeslie