When the AI ​​algorithms manage our individualized news rating, we get all the water on our mill until polarization causes democracy to crack. The question is how to solve the dilemma when technology is about to put the free word in conflict with the free will.

In the early 2000s, US President George W was named Bush. Like many Swedes, I was not a big fan, but one thing impressed. He gave speeches that appealed to the Christian right and secular voters at the same time. How was that possible?

The trick was simple. By weaving in Bible quotes, but using only unusual verses, the words could pass unnoticed past non-believers, while Christian voters perceived him as a deep believer.

As late as the beginning of the 2000s, these were tricks. But now we live in another world. We do not have to turn in and out to say two different things at the same time. Today, we have technical platforms that make it easy.

For democracy to work, individuals are required to be free to form an independent view of their views.

As long as citizens, through their freedom of speech, can stay informed and through elections are given the opportunity to express their free will, we will have a democratic government.

But as our scientific progress progresses, we are getting better at understanding how people make their decisions. In his latest book, author Yuval Noah Harari draws 21 thoughts on the 21st century example to its head.

If you imagine a future where we have become so good at manipulating people's decisions that we can decide what someone wants to think, then the circle is closed in the worst possible way.

Dictatorship and democracy merge when an absolute dictator alone can decide what others want, and then let them do it - democratically.

It may sound remote, but we are already well on our way there. It is tempting to believe that the polarization we have seen in politics is due to abuse of our social media a la Cambridge Analytica, but unfortunately the problem goes deeper than a sly political campaign.

It lies with the algorithms that automatically optimize themselves to serve your personalized news rating. If you want to see them in action, it is enough to watch a couple of movies that the earth is flat on Youtube for everything to suddenly appear to be about it or other conspiracies.

The trick with the algorithms is that neither they themselves nor their creators have a political agenda. They only focus on one thing - to maximize viewer time. The most effective has been to select content that amuses us, scares us or makes us indignant.

I think the accelerating polarization we see is the result of all people now wandering around in an individualized kaleidoscope of things that amuse and frighten them rather than participate in a common public conversation.

When our main information channels - without human intervention - reshape themselves, we are all caught by our own confirmation bias. Environmentalists are stunned by ignorant participants in the petrol tax revolt, while SD voters are becoming increasingly convinced of the imminent cultural collapse.

This is the result of leaving control of an AI algorithm that blindly maximizes profitability.

You can think of what you want, but it is nevertheless harmless compared to whether the algorithm's owner, as in Yuval Noah Harari's example, optimized the outcome on something else. Google and Facebook have a terrible power to imagine in the hands of a totalitarian regime.

And at the same time, it's probably just as bad to go into formulating a framework for how social media gets to choose its content. To regulate what one can say to people is by definition an infringement of freedom of speech, which is also a prerequisite for democracy. This is one of many fields where the development of artificial intelligence leads to a difficult-to-solve dilemma.

Technology is in conflict with the free will. The question is how democracy will survive if in the future we are forced to choose between its two basic conditions.