According to Scout.ai, a publication of journalists, professors, and science fiction authors, the answer is a resounding YES.
It is even possible that Donald Trump used data analytics to determine which cities, and states to visit, as well as commanded his own personal botnet army that outnumbered Hillary Clinton’s 5 to 1.
Naturally, I was skeptical of everything this article claims, I even slept on this story before hitting publish. I’ve since looked into their claims and they appear to be telling the truth about the technology and people involved. What is unknown is the extent and success of what Trump and his team accomplished, but given the fact Trump won the election it is entirely possible these technologies could have been a deciding factor. Which means artificial intelligence, machine learning, echo chambers, and weaponized propaganda just influenced the election—not the Russians.
Steve Bannon is on the board of directors of a privately held company out of London called Cambridge Analytica. They’re a data mining, analytics, and strategic communications company funded by Robert Mercer, a friend and supporter of now President Donald Trump. This company has developed strategic processes and products for automating the ability to track and analyze everyone’s social media presence, and then target and optimize advertising specifically to millions of people using machine learning to show them “news” that matches their political identity. They are then able to use the engagement data to geographically see where Trump could potentially persuade people to vote for him, especially in swing states. Or just poison the well, and persuade people not to vote for Hillary Clinton.
In other words, Steve Bannon and company figured out that social networks such as Twitter and Facebook put everyone into echo chambers, which I wrote about the night of the election, and realized they could use that to their advantage to persuade the masses into voting for Donald Trump. By using machine learning, which enables computers to learn without being programmed, they were able to develop software to analyze everyone’s personalities based on purchased data and Facebook. They then used that analysis to display specifically tailored ads or “news” on Facebook and other websites (sometimes following users across the Internet), showing these users dark posts that only that person saw. By doing this, they could significantly optimize engagement and target people they knew statistically had a high probability of becoming a Trump supporter. They could also target Hillary Clinton supporters, and potentially persuade them not to vote.
Those who did not engage, Bannon and company then knew to ignore and no longer pursue (saving time and money). Those who did engage were put into new echo chambers controlled by Steve Bannon and Cambridge Analytica. By continuing to advertise to them what the machine learning software determined they wanted to read, they were able to persuade people to vote for Donald Trump who may not have normally voted; or perhaps discourage people from voting for Hillary.
If all of this is true, Steve Bannon essentially used Silicon Valley’s own means of social engineering, echo chambers, and monetization strategies against those company’s and Hillary Clinton to persuade Twitter and Facebook users to vote for Donald Trump instead. Only spending a fraction of what Hillary spent, and with much greater efficiency and accuracy.
In addition, this article claims Trump also had his own army on social media similar to David Brock’s Correct The Record, which was spreading propaganda for Hillary Clinton. However, instead of having paid human trolls doing their dirty work like CTR, who didn’t actually believe what they were being paid to write about, Trump used intelligent robots and leveraged existing social media botnets. This way Trump, Bannon, and his team were able to control the exact messages and narratives they wanted to send, and amplify anything that fit their narrative. It is even possible they paid to use existing Russian or Ukrainian botnets they purchased or leased from stateless rogue hackers (which is consistent with the Russian hacking narrative but a completely new way of looking at the evidence).
They didn’t even necessarily have to create the content, merely amplify anything that supported an idea or concept they liked that would either influence people to vote for Donald Trump, or persuade them not to vote for Hillary Clinton. It probably didn’t hurt that Bannon was able to leverage his position with Breitbart News as well.
It would not surprise me if my own article from election night, which received over 9 million reads in the days after the election, benefited from being retweeted and promoted by this Trump botnet. I was promoting the story heavily on Twitter using hashtags and keywords the bots were potentially amplifying. I don’t believe my reads or views were faked because most of my traffic came from Facebook and Twitter, and I was also the top story on Medium for a few days.
However, it is still possible the story was amplified by these bots causing real people to actually read my article, and even follow me. The intention of Trump’s botnet was actual human engagement as compared to Hillary’s CTR spam like techniques. The article gained so much traction so quickly that even one of Trump’s advisors read it.
While I can’t prove any of this, the fact it is even a possibility is both equally exciting, and frightening. While it is possible my article really did spread organically and was not amplified by a Trump robot army, the possibility that I may have accidentally tapped into something few people knew existed is something out of a science fiction novel.
The even scarier thing about all of this is the technology outlined in this article does in fact exist, and everything I’ve proposed is technologically possible and is exactly what Cambridge Analytica claims they can do.
Disclaimer: This article is about technology and propaganda Trump might have used to win the election. It is not an endorsement of Trump or Steve Bannon.