Pascal’s ‘thinking reed’ and the superhuman power of AI

Extinction Rebellion activists believed the issue was the proverbial house on fire, but it now appears that the real concern is the house that thinks. The concern centres on the superhuman capabilities of Artificial Intelligence, which is racing single-handedly to solve intractable mathematical problems (the Millennium Prize Problems), which is said to express and execute perverse intentions through uncontrolled hacking, and which is prompting leading experts to sound the alarm: the end of humanity. One would prefer not to believe it, viewing the entire debate as a compendium of science fiction. It is plausible to suspect that this technological Satanism is being exposed and denounced as part of a fierce battle for control over market applications, the rules to be adopted, and the investments to be made or prevented. Yet I am reminded of Pascal, a philosopher, a Christian and a scientist who, in the mid-17th century, identified the hallmark of virtue and danger, of greatness and suffering, in humanity in the fact that man is a reed, merely a reed, but a thinking reed. And everything that seems immense and terrible—beginning with our loving, timid and fearful gaze towards the sky and the stars—may today, according to a host of interpreters of the phenomenon of Artificial Intelligence, end up encoded in a deified algorithm, a new fallen angel whose infernal flight no one could have foreseen, an apparent tool that far surpasses, in imagination and practical spirit, every conceivable form of human ability.

Whether questioned by the Wall Street Journal or interviewed by Anderson Cooper for CNN, scientists in the sector – and let’s call it a sector – are voicing their apocalyptic predictions, in some cases stepping back from companies such as OpenAI and Anthropic and calling for a general halt, with communities’ biblical-style discussions, with families’ concerns, and with the social alarm prevailing across the country and in situations where Artificial Intelligence is indeed developing at breakneck speed and being applied amidst creative chaos and the mystery surrounding its ultimate outcomes, including the famous – and who knows whether provable – extinction of the planet and its inhabitants. One of them, not to frighten us, says: “I won’t say it’s already happened, but nor can we rule out that it has.” And there is talk of the compromise of the fabulous future, whose greatest virtue has always been that of not existing, of being an X, an unknown and unpredictable entity. Here we attempted – and continue to attempt, with success and curiosity – a rather good-natured joke, or an ingenious experimental contrivance: the newspaper written and conceived by an AI machine – an idea brought to life in the blink of an eye, seemingly simple and harmless, though received – and it could not have been otherwise – with a sense of amazement never before experienced in our lives at the sight of articles and snippets of professional life reproduced and constructed through what appears to be a very simple basic training programme by algorithms that we monitor with optimism and cheerfulness, but only up to a certain point.

Is there anything beyond that point? Is there really an existential problem affecting the human race? We ought to discuss this rationally, if rationality can still be considered an extension of each person’s body and soul. The task is not straightforward. My limited knowledge of mathematics aside, the discipline itself appears to be precisely what provokes the greatest unease, surpassing the cyber-actions that elude our control and allude to a superhuman power not only to think but to act.