Because you don’t have to be afraid of ChatGPT. An experiment (ours!)

Will artificial intelligence produce “catastrophic effects”, as Musk puts it? Maybe. In the meantime, coexistence with newspapers tells us the opposite and demonstrates that AI is a super stress test for those who provide information. And that it can do well, under two conditions

It is possible that Elon Musk’s concerns about the state of artificial intelligence have a basis in reality and it is possible that the appeal with which the CEO of Tesla invited engineers from all over the world to pause the development of ChatGPT for at least six months, to avoid “potentially catastrophic effects on society”, may contain a grain of truth and it is difficult to blame Musk when he says that “powerful artificial intelligence systems should be developed only when we are sure that their effects will be positive and their risks will be manageable”. However, what we feel we can affirm without too much hesitation, with respect to the powerful development of ChatGPT, is that artificial intelligence, applied to journalism, does not produce any potentially catastrophic effect and in some cases, as we will try to demonstrate to you, is able to even produce positive effects, with perfectly manageable risks. Therefore, in the event that Elon Musk is interested in delving into the history of a newspaper determined to experiment artificial intelligence with dedication on its pages, we allow ourselves, without conflicts of interest, to suggest an interesting model to study: ours, of course ! A month ago, as you know, Il Foglio chose to conduct a decidedly reckless experiment: inserting a text written with ChatGPT every day in the pages of our newspaper and rewarding two readers at the end of each week with a subscription to the Il Foglio and a bottle of champagne. selected at random from all those able to identify the text written with artificial intelligence every day. We thought, initially, that it was an experiment to test the limits and potential of artificial intelligence, and so in part it was, but we realized that in the end the test in question was geared more towards the journalistic quality of natural brains, i.e. ours, and on the journalistic quality of artificial brains, i.e. those financed by Musk & co.