Artificial journalism preaches well and raids badly

An intractable identity instinct prompted me to be more demanding with the chat. So I asked who I am, who is Marco Demarco. And the answer was that I haven’t been since 2013. So I hastened to write a post on Facebook, perhaps to confirm to myself that I was still there. And I said that, before the fatal news, the machine rewarded me with prestigious prizes that I never received, like a Saint Vincent for my career; and with unfortunately unfounded merits, such as having written highly successful books, even one entitled Gomorrah: a journey into the economic empire and the camorra’s dream of domination. A case of homonymy? No. Saviano Saviano. Furthermore, I knew that Chat was not a mere search engine like Google and I had taken care to add some biographical details next to my name, useful to enrich and better articulate the answer. But here’s a double surprise: the machine reported, yes, and I would say slyly, some details that I had offered it, for example that I am a Neapolitan journalist. But it carelessly ignored others, like the books I’ve actually written. I will be told that I am presumptuous and that the machine cannot take everything into account, even trifles and pinzillacchere. But I was struck, among many others, by a message sent to me by Giulia Galasso. If it’s any consolation – she wrote to me – I asked about my father and he replied that he has been the CEO of telecom since 2004!. Yes. I later learned that, when asked again and better addressed, the machine only partially corrected the errors: Professor Galasso, for example, continued to appear as a senator despite having always been elected to the Chamber of Deputies.

So, still unsolved, is there also a problem concerning the so-called machine learning, the self-learning process of artificial intelligence? Evidently, yes. Which confirms another aspect of this complex affair. In other words, the emphasis with which technological innovation is too often written, describing marvels that are not yet such and focusing too much on the key of exceptionalism. And, consequently, the opposite excess, that of gradualism, of a knowledge, that is, which claims to have already seen them all.

He explains why I agree with the conclusions reached on these pages by Antonio Pescap: We are only at the dawn of a trial and we can know nothing about what will happen. Nick Bostrom, one of the leading experts on the subject, also said so: almost a century ago, it was already thought that superintelligence was around the corner, but that corner is still not seen. And if true about a superintelligence capable of annihilating human intelligence, even more so in the specifics of journalism. Further proof? An hour after the official election of Elly Schlein I asked ChatGpt if he could write a comment on the election of the new Pd secretary. Answer: Sure, as an AI I don’t have a personal opinion, but I can offer a neutral perspective…. And after the premise, this: The election of Enrico Letta as new secretary of the Democratic Party represents an important choice for the party…. Letta, still him. Correctly, the chat warns from the outset that on the most recent events it could have some failure, some updating problem. But also on the most important events? Even on those who most mobilize public opinion? So let me point out what happened at the School of Journalism last December 22, that is, two months before the primary school. There was a lectio magistralis by Paolo Mieli. And finally, the questions came. Including this one. Who will win the challenge in the Democratic Party? Mieli has no doubts, and his answer is documented by the report published on Inchiostronline, the school’s website. Elly will win them, Schlein, she says. And an explanation follows, while Bonaccini still seemed to have no rivals. By this I mean that the machine does not know and man even foresees? No, I won’t go that far. I’m just saying that a good analysis of the facts, which takes polls into account, but also much more, can help to better understand the future, even before it officially manifests itself. Speaking of newspapers and news like horses, however, I went to re-read what Bostrom writes about it. He says that after declining to 2 million in the early 1950s, the horse population in the United States has rebounded sharply to over 10 million. And he explains: Economic growth has allowed a greater number of citizens to indulge their passion for horse riding. Which coincidentally rhymes with information. Which reveals the metaphor.

The Corriere del Mezzogiorno newsletter

If you want to stay updated on the news of Campania, subscribe for free to the Corriere del Mezzogiorno newsletter. Arrives daily straight to your inbox at 12 noon. Just click here.