Last month, a group of rabbis walked up a winding hill in the heart of the Vatican, past the marbled dome of St Peter’s Basilica. They approached Casina Pio IV, a 16th century patrician villa, which now houses the Pontifical Academy of Sciences. Just a few metres behind, on a road lined with palm trees and cacti given by churches all over the world, came a group of Muslim imams and academics from Abu Dhabi and Berkeley, California. I followed, a witness to this rare meeting of minds. They were here to discuss an important matter with the Pope: the future of artificial intelligence.
Inside the villa, in a square golden room frescoed by some of the greatest Renaissance and Baroque artists, sat leaders representing the three Abrahamic religions. They were united by their concerns.
The meeting was years in the making, but it occurred just as the rest of the world had woken up to the potential of AI, due to the viral success of internet chatbot ChatGPT. The program is deceptively simple: a technology that can answer a user’s questions using natural language. ChatGPT answers can encompass everything from MBA exam solutions, business plans, and feminist essays on Frankenstein, written in styles and voices as varied as Shakespearean, pirate or 90s hip-hop.
The summit was called to discuss the broad umbrella of artificial intelligence, including decision-making systems, facial recognition and deepfakes. Soon after the launch of ChatGPT late last year, the pitfalls of a system mimicking human communication became obvious: it can spew falsehoods and misinformation, magnify societal biases in the texts it produces, and if you work around its guardrails, as users have already done, it can generate unsavoury and illicit content. All this can be achieved instantly and at scale, reaching millions.
The religious leaders agreed to convene in Rome because of what they see as an urgent need to make AI into a technology that respects the human world’s ethical and moral limits.
Before meeting with the Pope, three of the leaders — Archbishop Vincenzo Paglia, Chief Rabbi Eliezer Simha Weisz of the Council of the Chief Rabbinate of Israel, and Sheikh Abdallah bin Bayyah of the UAE, recognised as one of the greatest living scholars on Islamic jurisprudence — articulated their worries. The sheikh feared societal division due to misinformation, and threats to human dignity because of Big Data’s problems with privacy. The archbishop spoke of AI being used to curtail the freedom of refugees, through automated borders; Rabbi Weisz worried that we would forget that intelligence alone is not what makes us human. Unlike identikit technological tools, each of us is unique, he said, animated by what the devout might call a soul, which sets us apart from “any Man-made instrument”.
On a video link from Abu Dhabi, Sheikh bin Bayyah asked who we should hold responsible for the mistakes of artificial intelligence. What will happen to communication between humanity? How will artificial intelligence affect our behaviour? How do we avoid failures? Would we, he asked poetically, like the little silkworm, suffocate inside our own creation? Leaders from the technology world, including Microsoft’s president Brad Smith and IBM’s research chief Dario Gil, listened carefully before responding. AI was about to explode into public consciousness, and they expected it, in some sense, to transcend the human experience — but people must remain at the centre of its design.
Towards the end of the morning, the three figureheads — the elder Sheikh bin Bayyah represented in person by his son — signed a joint covenant alongside the technology companies, known as the Rome Call. The charter proposes six ethical principles that all AI designers should live by, including making AI systems explainable, inclusive, unbiased, reproducible and requiring a human to always take responsibility for an AI-facilitated decision.
As they signed the pact, the room fell silent. I looked at those around me and wondered if they, like me, were committing this moment to memory. Whatever disputes raged beyond this room, we agreed, at least for now, that the rights of humans in an AI-infused world were worth protecting.