Hello,
Thanks to those who answered my call last week for help ranking some courses I plan to create. And now, to quote King of memes Bernie Sanders, I am once again asking for your support. I’m doing some work around 'the data skills gap', i.e. when people/companies feel they don’t have the necessary skills to thrive in an increasingly data-driven world. Is this something you recognise? And if so, what are the sorts of things you would really like to be able to do with data skills? I’ve put a free-text box in the usual place, or feel free to email. For those watching events in Afghanistan with horror, I’ve had a scheme called Hostnation recommended to me. It provides friendship to refugees arriving in the UK. I can’t (yet) vouch for it personally, but it strikes me as a really good idea. Until next time, Oliver
Thought for the Week: Inside-Out World
Why is Shakespeare's work so well-known? Is it because he, more than other artists, wonderfully captures universal human themes? Or is it because he was popular with Queen Elizabeth, had friends willing to print his plays, and became an emblem of ‘Englishness’ for the Victorians to export to America and the Empire? In brief: should we look to his individual talent, or wider circumstances? Obviously we could just say 'a bit of both'. But let's try harder: is it 50:50, or skewed? As a sociologist – and also as someone who’s performed quite a bit of Shakespeare – I strongly lean towards circumstances. This isn’t me trying to be all contrarian, ‘Shakespeare isn’t that good actually, he got lucky’. I just think the idea that he captured humanity so much better than other artists, and basically everyone through hundreds of years agrees on that, is a hard one to stand up. In their ‘Ten Commandments for Aspiring Superforecasters’, Philip Tetlock and Dan Gardner argue that good predictions need to balance ‘inside’ and ‘outside’ factors. Inside factors are specific to the person or incident in question; outside factors are broader, repeated circumstances. Inside factors are often more appealing, and can therefore suck in disproportionate attention. But outside factors might explain more. A good example is the election of Donald Trump. His personality was such a dominant part of discussion. But this distracted from ‘the fundamentals’ – how would one expect a generic candidate of a challenger party to perform after the incumbent party has been in power for two terms (maybe also drawing on historical data about how challengers perform when the economy is in a certain state etc.). Looking at these, Trump had a decent chance. Such questions about the power of individuals vs. broader structures are popular amongst social scientists. But I recently saw an interesting parallel in more mathematical sciences. I’ve just finished The Physics of Wall Street by James Owen Weatherall (a well-written book jammed with fascinating content, though with an annoying low-level science bro tone* throughout). One of the many characters we meet in the book is French geophysicist Didier Sornette, studying Kevlar. Kevlar is a material used in many applications, including space flight. Sometimes small ruptures in Kevlar stay small; sometimes they blossom and fuel tankers explode. The problem – why do some ruptures stay small, and some grow very big? Looking at the ruptures, it seemed pretty random. But Sornette took the outside view. He examined the wider material structure of the Kevlar; he noticed that sometimes it was arranged such that any small ruptures would spread. The issue wasn't the rupture, it was the pre-existing structure. (Wetherall uses the metaphor of an industrial dispute – one random strike might not have an effect, but if there’s a strong underlying union structure in place then one strike can multiply up into widespread action). A key insight from this was that you can test for such structures by seeing how the material (or workforce, or whatever) behaves prior to the big rupture. This turns out to apply well beyond Kevlar. One example: by tracking the tremors of smaller earthquakes, you can spot when a particular area is ‘ready’ for a much bigger earthquake – and even predict when it might happen. A second example: Sornette then applied similar pattern-spotting methods to financial data. Doing so, he predicted an upcoming market crash in late October 1997. On 27 October, the New York Stock Exchange lost $650bn and was forced to close early for the first time in its history. This is not to say that markets are always predictable. This is a huge area of debate - unsurprisingly given the money is involved - and I’m not going to go into it here. But this is another example of what has been called “the unreasonable effectiveness of mathematics”, where a huge number of natural and social events (potentially even the outbreak of wars) seem to follow unnervingly precise mathematical rules. Let’s think again of our inside and outside. Such mathematical regularity implies that much of the ‘inside’ is actually an illusion. Individuals make all sorts of financial decisions (or political actions, or whatever). All have their own human uniqueness and distinctness. But after all that, everything ends up following the same old mathematical patterns. It isn’t quite as bad as the time that 19th century mathematician Pierre Simon Laplace argued that all the future was already set in place by the laws of physics, so bye-bye free will. (Chaos theory and quantum physics later put paid to that theory, thankfully). But it’s still a bit depressing to think that individuality might have so little impact. That's not to say every event can be mathematically predicted. These methods focus on biiig events, huge earthquakes and financial meltdowns. Those are very impactful on human lives. But so are relationships, day-to-day conversations, and all those other things that elude maths. But nonetheless it’s a reminder, depressing but important: sometimes if we want to understand society, the human story might be distracting rather than enlightening. * ‘Science bro’ was a relatively popular Twitter term when I was doing my PhD. I use it to mean a combination of (i) ascribing pretty much all positive qualities to ‘science’, such that pretty general critical thinking becomes “thinking like a physicist” and (ii) not reflecting that habits like using “he” as a default pronoun – “a good scientist will always check his results” – might not make your field terribly welcoming to over 50% of the population.
Fun Fact about: Internet codes
‘Top Level Domains’ (TLDs) are those bits at the end of web addresses - .com, .net, .org, and the like. Many of you will be aware that different countries have their own variants - .co.uk for the UK, .fr for France, etc. You may also, like me, have seen an apparent profusion of new TLDs. I’ve seen plenty of tech companies using .ai, presumably to make everyone think of Artificial Intelligence. I’ve also seen people using .me at the end of email addresses. I assumed these were new TLDs being created to break the professional monopoly of .com. But no! As I learned from this thread, they are a clever repurposing of country codes. .ai is Anguilla; .me is Montenegro. As with much else in hipster tech this was foreseen in Charlie Brooker and Chris Morris’s dark but brilliant sitcom Nathan Barley, in which the superhumanly crass lead character registers his website in the Cook Islands so he can end the address with .co.ck. This was in 2005.
Recommendations
Fun economics podcast: The Planet Money podcast, and its daily spin-off The Indicator, are brilliant examples of using the audio format in a really fun way. Plus the stories always makes me go ‘how on earth did they think of that?’. In a recent episode, to explain the bond market they actually buy a bond in a failing oil company called Hornbeck, christen it ‘Becky with the Good Yield’, and manage to make me feel weirdly attached to… a bond.
Language learning: The ‘Coffee Break’ podcasts (Coffee Break German, Coffee Break French, etc.) are an easy way to help free-time language learning. The hosts manage to capture the confusions and feelings of achievement very effectively.
Biceps: A recommendation for those like me who (i) don’t want arms like chicken legs but also (ii) don’t want to go to the gym and also (iii) don’t want to own any more heavy objects. Brilliantly simple solution – resistance bands, which use tension instead of weight. Here's an exercise you can do with them (other videos on site also recommended).
And finally, Food: You can tell that some form of normality is coming back by the increasingly late evenings I am spending in central London. For anyone around central London looking for excellent food, I can strongly recommend Mele e Pere (Italian) and Wun’s Tea Room (also superb for Soho-watching).
Comments