Will AI push western governments to censor social media companies like in China?

Do you recall those days when media channels were limited, and we stayed at home to watch our favorite tv-series? That time when in Europe the range consisted of one national and, if you were lucky, maybe a few foreign or channels? It was a period where the media was segregated, organized around the different political party lines. Today, such an approach seems unthinkable in the west. We’ve come to a point where we are struggling with too much choice, in a ruthless open landscape.

Good thing that companies like Netflix or Amazon help us make these ‘difficult’ choices in a sea of content. The media sector’s investments in artificial intelligence will only increase over the years to come because they realize that ever more relevant and personalized offerings is how you really bring home the bacon.

The biggest media company you’ve never heard of

The online company that has fully embraced the future of the new media is the Chinese company Toutiao (头条), which means ‘headlines’ in Chinese. It’s a news recommendation app that deploys AI to quench the consumers’ thirst for news and entertainment of all its readers. You may not have heard about it yet. But it’s Big, and not just in China. It’s the favorite of Silicon Valley VCs and was valued at 75 billion USD end of 2018. Buzzfeed, their American competitor, decided early 2018 to work with Toutiao to distribute its content in China, because “If you can’t beat then, join them”. In 2017 there was a rumor that Baidu (China’s Google) would buy Toutiao, but the management of Toutiao reacted jokingly that they must have swapped buyer and seller. Safe to say that it’s probably the biggest media company you’ve never heard of.

But how did it become so big? Simple. They pulled every trick in the AI book to make their algorithms highly addictive: they learn individual user behavior and update themselves in real-time for a more personalized and relevant reading experience. And in contrast to most other (social) media, Toutiao uses a dual dataset environment. Beyond what you read, share or like, their algorithm recommends primarily on the basis of what you write. In doing so, it pierces one layer deeper into your subconsciousness. Now, in the West, that (recommendations on the basis of what one writes, I mean) might not have such a big effect, but in China, the impact is significant. Let me explain.

Considering the government-driven censorship, one could very well assume that China’s internet is a dull and very limited environment, but quite the opposite is true. Chinese netizens create way more content - writing and posting pictures, blogs, video, streaming, reviews, ratings, etc. - than we do in the west. As far back as 2012 a staggering 76% of Chinese netizens created online content compared to a mere 23% in Europe. We are seeking information, while Chinese want to be visible online. With China’s strict media control at play, the internet became the most powerful tool for 1 billion Chinese in the quest to seek an identity of their own. Paradoxically, the west considers China’s regime as a limitation in human liberties, while this censorship seemingly contradictorily sparked an explosion in creative exploration for most Chinese netizens. It’s the same type of inventive dynamic we notice in kids that were grounded by parents. China is creating a vast amount of online content, its consumers are thirsty for information as inspiration, and the government is delegating the responsibility of content distribution to the online platforms like Toutiao. This means that, while the government decides what is acceptable and what not, it’s up to the media platforms to make sure that all content complies to those guidelines.

Toutiao is a perfect example of how China 3.0 is emerging as a consumer driven society: a consumer seeking the comfort of an algorithm that decides on his behalf what he wants to see, as long as the delivery stays relevant. Toutiao's AI algorithms predict with great accuracy whether certain content will be successful and go viral or not. As such, they are migrating us all from a digital age into a semantic era - a world of context. Beyond text, images are parsed through an object- and pattern recognition software, adding personalized meta-data to every image. Imagine that you are watching a football game for which Toutiao is able to determine your favorite player based on your viewing behavior, it could zoom into him live; display for you the exact images you favor whilst delivering a personalized commentary and relevant captions – all with their AI-technology. No wonder that the Toutiao app reaches an average viewer’s daily attention span of 74 minutes as opposed to 50 minutes for Facebook and 32 minutes for Instagram or Twitter. Two years old, and Toutiao was already profitable and is amongst the world’s fastest growing applications.

Relevance versus freedom

As we extrapolate Toutiao’s success to the future, worldwide, one can expect that every investment in AI in the media industry – anywhere in the world - will result in ever more relevant and personalized information. And this will have a major impact on information and other content consuming behavior. We will all rely on these super-algorithms to satisfy and confirm what we think and like. But what does that mean for control? Are we in charge, or are the algorithms? Don’t they know our preferences better than we do ourselves as they can cross-reference with millions of others who behave like we do?

Social media companies of tomorrow could deeply limit and impact our choices and behavior, similar to how the west views China’s media today. But how will the western social media platforms balance profit versus societal and individual goals? How will we keep our freedom in a world where we can no longer choose what we see, especially when we always like what we see? Is the solution - in order to avoid massive commercial or other propaganda - to limit these AI-algorithms beforehand? Shouldn’t we curate them before they censor us? Will this curator be an authoritarian father like in China or a permissive mother like in the west? Should it be the government? And if so, will this make us less of a democracy? Will AI push western governments to censor social media companies like in China in the end?