Wether it was Ronald Reagan (Pruitt 2018) or Bruce Springsteen and the E Street Band (Kirschbaum 2016) who ended the cold war, for those who were not around in the 1980s it is difficult to imagine the optimism that prevailed after the fall of the Berlin Wall on November 9, 1989, and the (mostly) peaceful dissolution of the Soviet Union. This optimism was expressed in Francis Fukuyama’s “The End of History and the Last Man” (1992), which claimed that Western liberal democracy had seen its final triumph and was here to stay.
But while 9/11 in 1989 ended the cold war, the other 9/11 in 2001 ended the decade of optimism. Liberal democracy has been on the decline in the new millennium and populism and authoritarianism on the rise. The reasons that are given for this reversal are various, most frequently post-11 political failures and rising economic inequality. Another frequently cited reason, however, are the dramatic advances in Artificial Intelligence (AI) since 2011 and especially the rise of content-selection algorithms in social media.
As Jason Lanier (2018) and Stuart Russel (2019), both AI insiders, point out, the problem with such algorithms, and the reason they are so successful, is that they do not simply learn what a user likes and then present the user with such items. Rather, they change the user’s preferences to make them more predictable. Like any rational entity, such algorithms learn how to modify the state of their environment, which in this case is the mind of the user.
Both Lanier and Russel claim that such algorithms played a prominent role in the rise of political tribalism and authoritarianism. Although Russel’s main message is that we need to have a plan for the possibility that AI will succeed, he puts the claim most forcefully. About the content-selection algorithms used in social media he states: “The consequences include the resurgence of fascism, the dissolution of the social contract that underpins democracies around the world, and potentially the end of the European Union and Nato. Not bad for a few lines of code … Now imagine what a really intelligent algorithm would be able to do.”
In the current issue of the American Economic Review, Ro’ee Levy (2021) has published the results of a large field experiment to estimate the effects of social media news exposure. His results confirm that social media algorithms limit exposure to counter-attitudinal news and thus increase polarization. However, he also seems to be the first one to present evidence that exposure to counter-attitudinal news is effective in decreasing affective polarization. Exposing people to other people’s views decreases polarization.
When I did my PhD in economics at the University of Zurich in the 1990s, my thesis advisor – the late Franz Ritzmann – gave me a piece of advice that I have tried to follow ever since, even though it requires a bit of mental effort. “Don’t read stuff you agree with. You won’t learn anything new” he told me. “Read stuff you disagree with, if you want to learn something.” Levy’s research seems to suggest that Facebook could do some good for the world by incorporating a little of Ritzman’s advice into its algorithms.
REFERENCES:
- Francis Fukuyama (1992). The End of History and the Last Man ascendancy of Western liberal democracy.
- Erik Kirschbaum (2016). Rocking the Wall. Bruce Springsteen.
- Jaron Lanier (2018). Ten Arguments for Deleting Your Social Media Accounts Right Now.
- Ro’ee Levy (2021). “Social Media, News Consumption, and Polarization: Evidence from a Field Experiment.” American Economic Review, 111 (3): 831-70.
- Sarah Pruitt (2018). The Myth That Reagan Ended the Cold War With a Single Speech https://www.history.com/news/ronald-reagan-tear-down-this-wall-speech-berlin-gorbachev
- Stuart Russell (2019). Human Compatible Artificial Intelligence and the Problem of Control.