Summer of Reading

What is the summer for, if not reading? Here is a small summary of the things I have read this summer sat in parks, lying on grass, in bed and in airports and on trains. In no particular order, with a scandalously low proportion of fiction and purposefully avoiding direct links to Amazon where possible, I present my summer of reading.


Panama Papers

A slightly underwhelming account from one of the journalists at Süddeutsche Zeitung of publishing a huge leak of documents from Panamanian offshore financial services firm Mossack Fonseca; better known as the Panama Papers published in 2015. The mechanics of receiving a huge unstructured data dump anonymously are interesting, as well as organisation of a large international team to collaborate together on it. But the most illuminating aspect was the extent to which computational methods were required to uncover this information that is purposefully obfuscated, from unstructured scans of documents relating to shell companies and nominee directors. While the reconstruction of networks of related companies and simple fuzzy matches of names that are challenging in their numerous transliterations are data munging 101, these skills are not commonplace among journalists. Or at least for now.

Some individual tales of corruption are told from all corners of the globe, Russia, Venezuela, Germany and beyond. Yet if you are not already familiar with the actors involved these are unlikely to seem worthy of the pages and pages of detail devoted to them. Overall, Nicholas Shaxton’s Treasure Islands is a much safer bet to learn the jaw dropping extent of offshore banking and the brazenness of those making use of it. The authors seem to tacitly acknowledge that fact by means of frequent references and quotes.

Twitter and Teargas

Zeynep Tufekci is one of the more outspoken commentators on the intersection of technology and society. Her account of protest movements is buoyed tremendously by her long standing association with movements such as the Zapatistas going back several decades. Her main contention is that social media tools have lowered the barrier in the initial stages of recruiting to a movement, but this simply defers the tricky matter of agreeing upon principles and tactics that can cause that movement to later implode. Tufekci is insightful in recognising that the rough and tumble of heated arguments at relatively private hustings and AGMs have now been replaced by vicious Twitter threads and sharing of deleted screenshots among online communities striving for consensus. Yet all things considered, Tufekci is more optimistic about the ability of new technologies to increase and enhance civic engagement than other commentators such as Evgeny Morozov

Tufecki’s well established credentials in protest movements lead her book to age better, providing more general lessons than the Naked Diplomat (see below) which already seem somewhat naive and redundant less than 12 months after being published. But despite having first hand experience of several notable protest movements (Tahrir Square, Gezi park, Occupy Wall St, Hong Kong democracy protests and the Indignados movement), I can’t help but wonder if such deep examination of the nuances of behaviours of the participants leads to a brittle overlearning of these examples. Surely the majority of protest movements; obscure, unsuccessful and short lived, should be put under the microscope.

A Line in the Sand

James Barr has emerged as one of the most knowledgeable historians of the tumultuous half century between the collapse of the Ottoman Empire to the wave of Arab nationalism in the Levant. This is a thrilling account of the infamous agreement between the French and the British to divide up the former Ottoman provinces in the Middle East into independent countries. No party emerges with their reputation intact, neither the British, the French or a myriad of other actors. This is complex series of crosses, double crosses and machinations that is all the more chilling with knowledge of present day Palestine, Iraq and Syria.

A diplomatic cable from the time recalls the sad fate of Sharif Hussein and exemplifies the jarring twists of fate in the Middle East. Hussein had been the Emir of Ottoman Mecca and was selected by T E Lawrence to be championed as leader of Transjordan following the Arab revolt and even briefly had a claim to Caliph of the entire Muslim community. Yet his star waned and his claim diminished until he found himself escorted to a deserted train platform on the Hijaz trainline, with only a couple of suitcases, to live out an exile in Cyprus.

As the contradictory model of nation states and self-determination continues to be questioned, this should be added to the everyone’s reading list under ‘Compulsory’.

The Social Animal

The Social Animal is a hard book to classify. In parts a popular science book, in others a very firmly grounded self-help book or an outcome oriented guide to raising a family. It is written in the style of a piece of fiction describing the life of Harold and Erica and the effects of their nature and their nurture; how their own decisions and upbringing affect the trajectories of their lives. The book draws upon numerous empirical findings from the social sciences about the effects of poverty, intelligence and psychological well being on success and happiness.

A Most Wanted Man

Translating a good book to film is rocky endeavour, rarely a satisfying experience if you consume book-before-film rather than film-before-book. Ever since I saw and fell in love with the Constant Gardner, I wondered if le Carre could pull it off again. This is simmering thriller from John le Carre about a young Chechen refugee who turns up on the doorstep of Turkish immigrants in Hamburg.

The Naked Diplomat

The former Ambassador to Lebanon (the UK’s youngest ever Ambassador indeed!) writes here on how diplomacy should modernise in the face of new technology.

A few things struck me. Firstly, I naively recall the thrill and confusion back in mid 2011, of the early Arab Spring. This was when Fletcher took up his post in Beirut and incidentally the same time I moved to Dubai. The medium du jour was Twitter, which was central to organising among protesters in Tahrir and Gezi and in the ascendancy. For a long time Twitter was my preferred medium and an amazing way to get inside the heads of some incredibly smart and experienced people. It was an era of ambitious scientific papers claiming to predict stock markets and other things using Twitter as a global population sensor. Sadly it is a shadow of what it once was as so many people I admire drop off, mostly women in the face of unchecked abuse. It is an uncomfortable bell-weather that the main medium to replace it is WhatsApp; a private and encrypted channel that is predominantly one-to-one communication replacing an uncurated (in it’s heyday) public broadcast channel.

Secondly, Fletcher talks a lot about ‘soft power’. The ability of Daniel Craig and Adelle to raise the UK’s profile and connect with his Lebanese constituents. It is easy to forget just how powerful these cultural icons are, particularly as we prepare to pitch ourselves to the rest of the non-European world with increasing urgency.

Finally, I am now more receptive to Fletcher’s controversial 2015 open letter to Lebanon as he left his post. He proposes a slightly playful, less buttoned up approach to diplomacy that is not afraid to gently ruffle feathers and provoke discussions. His case is definitely helped by his engagement with the ensuing debate featuring lots of accusations of neo-Orientalism; a sign that you might well be doing something right, in the Levant. After a side-splitting response from Lebanese satirist Karl Sharro Fletcher gamely responded on YouTube.

Prince of the Marshes

Rory Stewart was appointed as the highest ranking civilian administrator in the Marsh Arab region of Iraq following the invasion and toppling of Saddam Hussein in 2003. While most accounts of this period in print and on film, such as Life in the Emerald City, have been focused on Baghdad, this provides a fascinating window into life in Iraq’s far flung provinces. Stewart thankfully had only a single serious security incident to deal with spent most of his time trying to navigate the landscape of sheikhs, religious leaders and eccentric strongmen when establishing civic bodies and councils. Stewart refrains from criticizing the occupation too strongly and is pragmatic about the job to be done of requesting money from Baghdad and disbursing it effectively to build schools and roads. His historical awareness is refreshing, quoting Lawrence and Machiavelli but perhaps too often.

The Master Algorithm

The history of machine intelligence has a few key dates that anyone seeking to rationalise the recent debate and hubris around the Future of Jobs, Deep Learning and the meaninglessly nebulous ‘Algorithms’, should be familiar with. The first is the burst in progress in areas such as machine translation in the post-war and Cold war years, flush with government funding. The second is the Winter of AI, when the low hanging fruit withered on the vine to reveal the limitations of more theoretical methods closely associated with Chomsky. This period is notorious among computer scientists older than 30 years; that is to say the less quoted subset of the community. The current mania can retrospectively be seen as the end of the wait for hardware to catch up. The truth is that we are not living in a golden age of enlightenment, but simply applying incremental iterations on methods first introduced 20 years ago using newly affordable graphics processors that can fortuitously be re-purposed from gaming to number crunching. Moreover the final applications in our sights may be tremorous; self driving cars and flawless facial recognition, but are extremely narrowly defined.

From this perspective, the goal of Domingues’ The Master Algorithm is a welcome return to the ambition of the first golden age of AI when researchers truly grappled with the idea of a flexible and omnipotent machine intelligence. Domingues’ writing is brisk and lively covering good ground; from Hume and the British empiricist tradition to autoencoders and Boltzmann machines although some of the analogies employed can be confusing and hard to follow. Six distinct schools of learning are introduced with corresponding algorithms and refreshingly, the ‘connectionists’ (equivalent to neural network proponents) are given equal weight alongside other schools of thought.