Long Live the Digital Revolution?

22 August 2019

Posted by Paul Jourdan on 22 Aug 2019

On 26th June Amati AIM VCT held its AGM and following this, as is now customary, a series of investor presentations and the Amati Guildhall Creative Entrepreneurs’ Award. Last year I had referred to the work of the British-Venezuelan scholar Carlota Perez on the shape and phases of industrial revolutions as a way to offer an explanation as to why the ten years since the great financial crisis (of 2008) had been so fruitful for Investors who were backing companies riding the waves of the industrial revolution based on digital technology, which had began in the early 1970s with the invention of the microprocessor. The period from 2008 on could perhaps be seen as the “implementation“ phase of this 50 year industrial revolution, and this co-incides with what Perez calls the “Golden Age”. To bring home the astonishing potential of artificial intelligence I showed a move in a chess match between Stockfish 9, the most powerful conventional computer chess engine, and AlphaZero, a neural network based self-learning computer chess engine written by Google’s Deep Minds last year. In this move Alpha Zero sacrifices a piece where the compensation comes 19 astonishing moves later. This was a tale of optimism, tempered by the question of when this particular industrial revolution matures and morphs into something else.

This year, perhaps coloured by a more difficult period in markets, I suggested that these trends have matured significantly and that some dark stains have begun to appear on the hitherto almost universally admired character of the digital age. I suggested that the two dominant giants of personal data, namely Facebook and Google / You Tube (or Alphabet as its parent is called), both the embodiment of digital cool, had perhaps unwittingly become as much a threat to the values on which we build our society, as they are facilitators of cutting edge of the digitally aided life. As a perfect metaphor for this shift Google dropped its famous dictum “don’t be evil” from its code of conduct in mid-2018.

The problem I was drawing attention to has been eloquently described by Roger McNamee, long standing Silicon Valley investor, and one time adviser to Mark Zuckerberg in his book “Zucked: Waking Up to the Facebook Catastophe”. The problem he outlines is just as apparent with You Tube, and it relates to the way that both Facebook and You Tube (together with all of Google’s other digital platforms) capture vast amounts of information about users, whilst at the same time being trusted by them to deliver them content that is thrown into the mix of the things they were looking for, or into the flow of conversations with friends, and it does this based on calculations which maximise profitability.

McNamee puts it like this: Facebook “is a massive artificial intelligence that influences every aspect of user activity, whether political or otherwise. Even the smallest decisions at Facebook reverberate through the public square the company has created with implications for every person it touches”. He adds, “The fact that users are not conscious of Facebook’s influence magnifies the effect. If Facebook favours inflammatory campaigns, democracy suffers.” And it turns out that in the business models of both Facebook and You Tube profit is maximised by increasing consumer engagement, and the most powerful way to do that is to appeal to the so-called “Lizard brain” emotions of fear and anger, as these produce “a more uniform reaction and are more viral in a mass audience”.

In its 4th May edition the Economist carried an article entitled “The Tricky task of policing YouTube”, which touched on the same issue. “An algorithm and user interface engineered to maximise ‘watch time’ keeps users on the site in part by serving them progressively more extreme videos on whatever subject they happen upon – a ‘rabbit hole’ that can lead those curious about a global tragedy into conspiracy theories or rants by white nationalists. A senior executive said in 2017 that recommendations drive 70% of the site’s viewing. The site’s engagement-driven model in turn rewards those who provide more outrageous content”.

If we want to understand why it should be that the maturing of the Golden Age of the digital revolution should coincide with the emergence of a depressingly post-truth public sphere and dangerous resurgence of populist politics the business model and detailed workings of Facebook and Google/You Tube have to be a major factor in any explanation. Even if there were no so-called “bad actors” this would be enough to tilt the world towards angry and extreme politics. But the problem is made worse because since 2016 it has become painfully obvious that there are powerful “bad actors” who are largely unseen, who have magnified the effects. Cambridge Analytica was just one that happened to get caught, thanks to the diligent detective work of journalist Carole Cadwalladr. In her now famous Ted Talk, “Facebook’s role in Brexit – and the threat to democracy” she concludes that she has stumbled on an issue which is bigger than Brexit, bigger than Trump, bigger than Silicon Valley, and implores us to wake up to this threat before it consumes the freedoms that we have long come to take for granted.

The same extraordinary power of artificial intelligence which is beautiful to behold in Alpha Zero’s chess becomes something terrifying when applied to the science of manipulating public opinion via the new media platforms which the public treat as innocent play things, and are therefore to which they are still utterly susceptible. Those who think that Brexit and Trump (the first major consequences of this phenomenon in Western politics) are good things for the world might be tempted to dismiss this whole issue as the ravings of sore losers. This would be a profound mistake. This becomes clear by reflecting on its manifestation in other parts of the world, of which perhaps the most appalling example was the Rohingya massacres in Myanmar. The title of an article in the New York Times from 15 October 2018 sums up the gist, “A Genocide Incited on Facebook, with Posts from Myanmar’s Military”. Facebook itself has apologised for not doing more to prevent this genocide. The pressing question is this: how do we uphold confidence in the democratic process and rule of law with such a potent ability to manipulate public opinion in ways that are largely invisible being available for sale. We have enjoyed fantastic fruits of the digital revolution for the last few decades, now we have to come to terms with some extraordinary threats as well.