Candidate for artificial intelligence for mayor in Taranto tells us so much of our relationship with politics

An algorithm Prime MinisterThat is why the values ​​and programs of the winning implementation would be their own and would evaluate as policy rules based on them. As an alternative you could decide to vote on a series of specific issues (should Italy have to leave Europe? Should we close the ports? Do you have to increase the minimum wage?) And then Let it be an algorithm that determines the best way To complete these goals (which was more or less the way assumed by Lars leaderThe candidate chatbot for joke in the Danish elections of 2022)

Those who would certainly support such a thing is the founder of the 5 -star movement, Gianroberto Casaleggioaccording to which “An idea is not right or left. It’s a good idea or a bad idea”. Such a dichotome vision (or, if you want, binary) as naive. For example: reducing poverty is a good idea, every politician would agree on this. Do this Increasing taxes for the rich And the redistribution of the yield is a left road; Do this Reduce the taxes on the richIn the hope that this promotes economic growth and therefore employment, is a right -wing road.

Which of the two would opt for artificial intelligence? It all depends on the many boundaries in the form of bias, hallucinations, etc. – of what is included in the data with which it has been trained, that is, the solutions that have been tried in the past and the results obtained; Ignoring that a road that has been successful in a specific context can fail in another that the social -economic conditions can be changed, etc.

The boundaries of art policy

In the field of administration, the most important limits that collided, When testingArtificial intelligence are in principle three: Use data with regard to the past to model the future (with the risk of perpetuating prejudices and stereotypes), is in favor of the statistically dominant champion (punishing the minorities) and is often unable to change its analysis according to the different contexts.

But is it really that it should work an artificial intelligence in the service of politics? Most likely, no. Artificial intelligence is not a replacement for humansBut a little more looks like one of his assistant. It’s not against us, but it’s a supplement to us. We are better in analyzing the context and identifying unexpected solutions, the machine can offer us exceptional support Grind huge amounts of dataFinding correlations invisible to us.

For these reasons, artificial intelligence can be An extremely useful ally of the human decision makerBut can’t replace it. The fact that something like this introduces itself – even in a provocative context such as those who led to the “candidacy” of Anna Luce d’Amico – probably tells us very little about the real potential of artificial intelligence. Instead, he tells us a lot about the The distrust level that we are aged at politics.