Can artificial intelligence replace politicians? A quarter of French people say they would be willing to delegate government decisions to algorithms.
More and more jobs are being automated thanks to – or because of – artificial intelligence. Perhaps next on the list is a politician. In any case, if we are to believe the results of an amazing study published by a research center at the University of Madrid on 2,500 people in 8 European countries.
To the question: “Would you be willing to let the algorithms rule your politicians?” 25% of the French answered yes, 30% in Germany, 43% in the Netherlands. Real feeling or just an expression of fed up of politicians?
Artificial intelligence achieves 9% of the polls in a city in Japan
The idea is to develop computer programs that are able to make cold, objective decisions based on thousands of criteria, based on the public interest, without hesitation, without influence, and without reneging on their promises. No longer men but “political machines”, cold, rational, incorruptible. It sounds totally crazy, but some people think about it.
In 2018, during the municipal elections, the Japanese city of Tama, among the candidates was 150,000 residents, Amnesty International. On the posters, he took the form of a robot, with a “neutral and objective” policy implementation program. He was not elected but got 9%.
Simple decision help?
Except that AI cannot make decisions, it has no judgment and no moral meaning. Human judgment and morals cannot be computerized. In fact, AI will be useful, not to make a decision in the strict sense of the word, but to support the decision.
He is very good – much better than humans – at analyzing tons of data and drawing conclusions about the evolution of the situation or the public policies to be implemented. We can ask him questions like “How to improve the transportation network in a city or country” to “Should retirement be increased to 65?” Or even an optimum level of taxation and public spending so that the economy is in good shape.
It will take into account thousands of parameters, imagine all possible scenarios and give results, revealing patterns invisible even to a trained human eye. We used it at the time of the “Great Debate”, after the Yellow Vests, to analyze and synthesize hundreds of thousands of contributions from the French – often handwritten – and we used algorithms to try to make sense of words or sentences, and thus the prompts or propositions that appear most often, to bring them to the surface.
Automated speeches?
Algorithms can also be used to write speeches. In the United States, a researcher has even developed an algorithm capable of writing political speeches. A computer tool that we have fed with thousands of speeches and is able to create one from scratch on a specific topic. Finally, the cabinet ministerial function is more than that of a politician in the strict sense of the word. But she (fortunately) will not be able to decide what to do in the conflict between Russia and Ukraine. And they have no vision of their own, for the future of France for example.
These are companies that have already started delegating power to algorithms
A few years ago, a very serious venture capital firm in Hong Kong appointed IA to its board of directors, which has exactly the same voice, the same decision-making power as the other members of the board. Its name is lively, and its task is to swallow large amounts of data and analyze the pros and cons of investment decisions of this company in a completely objective and indifferent way, unlike humans, who are always biased.
Even if, in fact, this “neutrality” is highly debatable. The algorithm also always has biases, it is not made by the Holy Spirit, it is written by humans, with their strengths and weaknesses. There is an expression for this in computer science: we say “garbage in, rubbish out”. Basically, if you feed programs with garbage, they can only come out of the garbage.