AI and democracy

Democracy is endangered. We have long since crossed the line where digital technologies and social media began to affect not only our personal lives but also our political choices. Newspapers are vanishing in favour of social media networks, which now provide us with the latest news, while bots flood our feeds with vast amounts of fake information – or at least information that serves specific interests.

Social media platforms are designed to exploit our attention and our natural craving for small, instant gratifications – all with the goal of selling ads and products. Shoshana Zuboff described this dynamic in her book Surveillance Capitalism, and Lawrence Lessig refers to it as the “engagement business model”. AI is now making these timeline algorithms even more fine-grained. But can we be assured that this model truly serves the many – or just a few?

Another side effect of this attention-based model is the preference for certain types of content: hatred, violence, and extreme views are easier to track – and easier for AI to process. But just because it’s easier – can we really say it’s better for us?

People with more extreme political views are easier to target through algorithms. So would it be fair to say that it is in the interest of the company behind the algorithm to extremify political opinions?