Thank you very much, Mr. Chairman. It's a privilege and an honour to be here in front of this assembled international committee.
I appear before you this evening as an unlikely witness. I say that because I've spent pretty much my entire career promoting the virtues of the open Internet. I came of age during the Internet revolution of the late 1990s. I worked on the first truly digital political campaign for President Barack Obama in 2008. I was one of Hillary Clinton's digital diplomats in the heyday of Internet freedom during the Arab Spring. It was a moment in time when it seemed like smart phones and social media were the genuine catalysts of social and political movements to democratize the world. It was an inspiring moment. These technologies did help those things happen.
I'm an idealist at heart. I wanted to be in the middle of that revolution, but I sit before you today as a troubled idealist. I went back and worked for my old boss in 2016 on her presidential campaign. I ran the technology policy advisory committee. I had a ringside seat to what happened in America between 2015 and 2017. What I saw was that the open Internet that was meant to expand freedom instead turned into a powerful technology of social manipulation and political distortion. You all know the story. What was once the great hope for the revitalization of democracy is now considered by many to be among its greatest threats. My friends, that is a bitter irony—bitter—but it doesn't need to be that way.
The promise of information networks to distribute power to the people is a promise that we can reclaim, but we need to see at what point the astonishing control over wealth and power in this industry began to develop and steer things off course. The roots of this are deep, and we can track it back for decades, but I pinpoint a moment in time between 2014 and 2017 when machine learning technologies were applied to social media platforms, so-called artificial intelligence.
These technologies were not core to the Facebook and Google business models in 2011 and 2012 during the heyday of the Arab Spring. They arrived on the scene sometime later. If you want to know exactly when they arrived on the scene, look at the profit and revenue charts of Google and Facebook. I've written down the numbers for Facebook just to give you the case in point. In 2011, Facebook for the first time made $1 billion in profit on $4 billion in revenue. In 2017, just six years later, after the advent of these new technologies, they made $16 billion in profit on $40 billion in revenue. That's a more than 10x increase in six years.
How did that happen? It happened because they figured out a business model for superprofits. Step one: Track everything that billions of people do online and put it in a database. Step two: Sort that data and group people into target audiences and then sell access to their attention, engineering your entire information marketplace to optimize not for the quality of information or the civility of the dialogue in our society, but optimize just for addictiveness and time spent on the platform. Because the more time people spend on the platform, the more ads they see, and the more money the superprofits make.
It's a beautiful business model, and it works. It works with 10x profit in six years. Very few companies can claim anything like that kind of growth.
Also, it's not just the ads that get targeted. Everything gets targeted. The entire communications environment in which we live is now tailored by machine intelligence to hold our attention. This is not a recipe for truth and justice. What feels true performs better than what is true. Conspiracy and hate have become the organizing themes of social media, and that is a space that is easily exploited by propagandists peddling bigotry, social division and hatred to the disillusioned.
This is the connection between the data markets that we've heard talked about at this table and the abhorrent content you see online, whether we're talking about everyday hate speech or about something truly awful like the shootings in New Zealand. It is the algorithms that lead us into the temptation of our biases. This is what we have to address. We cannot rely on the industry to fix this problem.
The core problem lies at the heart of the business model, what Professor Zuboff calls “surveillance capitalism”, and these companies are kings of the market. Public traded companies—