By Eli Kerry
According to recent findings by investigative journalists at the Observer, a reclusive tech billionaire named Robert Mercer is one of the most influential figures behind both Brexit and the rise of Donald Trump.
Mercer made his fortune at IBM, where he was a brilliant computer scientist. His breakthroughs in language processing — a field critical for the development of A.I. — have been called “revolutionary” by the Association for Computational Linguistics. He has channelled the money he made there, as well as in hedge fund management, into such pursuits as a series of yachts all named Sea Owl, a $2.9 million model train set, climate change denial think tanks, and the manipulation of the mainstream media.
The largest example of Mercer’s interest in shaping politics is Breitbart, the website which acts as the internet’s alt-right headquarters. Explicitly established as a rightwing Huffington Post, the site launched the career of Milo Yiannopoulos and regularly plays host to Islamophobic and anti-semitic viewpoints. Steve Bannon, who left his executive position at Breitbart to become Trump’s White House Chief Strategist, funded Breitbart with $10 million from Robert Mercer, his close friend. Now, Breitbart has grown larger than its inspiration, the Huffington Post, and it is the biggest political site on both Facebook and Twitter.
Mercer also has a $10 million stake in a small data analytics company called Cambridge Analytica, which specializes in “election management strategies” and “messaging and information operations” refined over 25 years in such places as Afghanistan and Pakistan. These types of tactics are known in military circles as psychological operations, or “psyops.”
Cambridge Analytica closely follows Mercer’s lead. It reportedly worked for the Trump campaign as well as the Leave campaign in the United Kingdom. Steve Bannon’s presence also seems to follow from Mercer’s money: until recently he had a seat on the board of Cambridge Analytica.
According to Cambridge Analytica’s website, the company maintains psychological profiles on 220 million Americans, each one of which incorporates 5,000 pieces of data. Elon University’s Jonathan Albright, a professor of communications, says that this system amounts to a “propaganda machine” in which Cambridge Analytica can use trackers from sites such as Breitbart to follow visitors around the internet and then target them with ads on Facebook.
Cambridge Analytica has denied assisting the U.K.’s Leave campaign in any way. But according to Andy Wigmore, the Leave campaign’s communications director, Cambridge Analytica had worked for them, teaching them to build profiles, target people, and lift masses of data from people’s Facebook pages. Wigmore says Facebook, especially Facebook ‘likes,’ were key to the operation.
“Because using artificial intelligence, as we did, tells you all sorts of things about that individual and how to convince them with what sort of advert … The computer never stops learning and it never stops monitoring.”
But it was true that the Leave campaign had not formally employed Cambridge Analytica. Instead, Wigmore explained that the firm had worked for the campaign for free, thanks to Robert Mercer’s friendship with Nigel Farage, the British politician who orchestrated Brexit.
And Cambridge Analytica’s data surely proved useful. Not much is known about the company’s psychometric model, except that it originated with scientists at Cambridge University’s Psychometric Centre, and that this research was based on a Facebook personality quiz that went viral — over six million Facebook users took the quiz.
By correlating data like this with millions of other people’s information, the model can predict peoples’ actions. According to Michael Kosinski, the center’s lead head scientist, knowledge of 150 of a person’s ‘likes’ is enough for the model to predict someone’s personality better than their spouse. 300 likes and it can make better predictions about yourself than you can.
The internet is increasingly a battleground for mass propaganda in this vein. For example, according to Sam Woolley at the Oxford Internet Institute, just before Britain’s E.U. referendum, one-third of all traffic on Twitter consisted of automated “bots” — fake accounts programmed to look and talk like people, and shift national consensus by making certain topics trend. The bots — most of which are Russian — were overwhelmingly in favor of Brexit. Before the American presidential election, the same bots were pro-Trump.
The future of this social media weaponization is uncertain. But one thing that especially concerns Woolley is the hundreds of thousands of “sleeper” bots he’s found: bots which have only tweeted once or twice, and are now waiting for some kind of trigger where they will rise up and drown out other information.