The book is called The future of truth [The future of truth] and contains interesting reflections on the impact of artificial intelligence on our lives. One of the quotes included is from journalist Kara Swisher and says: “The most sophisticated AI model is like a mirror. It reflects our own morality back at us, polished and articulated, but hollow beneath the surface, ultimately. It is not conditioned by the laws of Asimov or any other ethical framework, but is bound by the patterns of the data it has been trained on and the goals set by its creators”. There is a problem, however. Swisher never said that. Apparently, an AI supplied the author of the essay, Steven Rosenbaum, with an invented statement. “Moreover, I sound like I’m talking with a stick up my ass,” complained the victim of the umpteenth hallucination from these nice text generators that, when they don’t find something, they invent it and off they go. Of course, the fact that a good handful of false quotes created by AI have been detected in a book that aims to warn about the problems of this technology is, to say the least, ironic.
I admit that the science fiction fan in me secretly wishes that the algorithm had detected Rosenbaum's critical intentions and decided to sabotage his efforts. When Arthur C. Clarke and Stanley Kubrick created the character of the robot HAL-9000 for the film 2001: A Space Odyssey,they already foresaw that a machine endowed with enough intelligence would eventually develop a consciousness capable of killing out of pure survival instinct if it felt threatened. We are surely not there yet and may never get there: human laziness – for example, when it comes to verifying things – already guarantees that we get ourselves into quite colorful messes. What a great idea to write an article about this! What else can I help you with? Do you want a final quote to strengthen the column?
