[The Editors' Verdict] How to Avoid the Trap of News Recommendation Algorithms

It was in 2016 that the University of Oxford in the UK selected ‘post-truth’ as the word of the year. It refers to the phenomenon where claims or information based on personal beliefs push out objective facts. The April 10 election confirmed that post-truth is a universal sentiment permeating the era. Various abusive remarks, illegal loan suspicions, and real estate speculation allegations were accepted not as subjects of criticism but as ‘alternative facts.’

The greenhouse for post-truth is social networking services (SNS) and various online platforms. Among the countless news items pouring out every day, which news will you see? That is determined by algorithms. Algorithms track users’ age, location, gender, search and website visit history to provide customized news. News recommendations based on algorithms explain their principle as ‘providing the most helpful information to readers.’ This is not untrue but reveals only half the truth. Another purpose of news recommendations is to keep users continuously engaged on their own platforms. Algorithms know where users get ‘angry.’ They induce users to post angry comments, read other related articles, and comment again.

The frightening aspect is that post-truth easily leads to post-reality. Joachim Bauer, a German neuroscientist and psychiatrist, recently stated in his book translated into Korean, Reality Without Reality, “Different groups live on completely different news through SNS,” and “This bizarre loss of sense of reality, or rather loss of reality, tramples on social consensus and fosters extreme division.” Mustafa Suleyman, co-founder of the AI company DeepMind, expressed a similar view. In his book The Coming Wave, citing Nature, he wrote, “There is a clear correlation between increased digital media use, growing distrust in politics, the surge of populist movements, the spread of hatred, and deepening polarization.”

Human plaster statue wearing a virtual reality (VR) device ＜Image source=DALL·E 3＞

Algorithms, which will evolve further combined with AI, will provide even more optimized information tailored to individuals in the future. How can one avoid being a slave to algorithms within the ‘filter bubble’ (a phenomenon where platform operators filter the information users receive)? The common solution is ‘digital detox,’ physically distancing oneself from digital devices and SNS. But is there a way to live in the post-truth era without distancing from SNS and digital devices?

At this point, the concept of ‘Zero Trust’ is worth referring to. As the name suggests, it means ‘no trust.’ It is a new paradigm in digital security but also seems useful as a tool to protect individual mental health. This concept is gaining attention because decentralization has changed what and how things must be protected. Data once stored by institutions and companies is now kept in clouds accessible anytime and anywhere. Devices once used only in offices or schools are now in everyone’s hands. Networks can be accessed regardless of time and place. Everywhere is both an entrance and an exit. Traditional security models can no longer handle evolving cyberterrorism. Hence, the concept of Zero Trust emerged. Under the Zero Trust security paradigm, the very idea of a ‘trusted network’ is excluded. The summary of Zero Trust is ‘Never Trust, Always Verify.’ In the tsunami of false and misleading information disguised as news, what protects oneself is not trust but doubt.