Skip to main content

Creativity and AI – A Report from the Vatican

AI media can be surprising, useful, clever, inaccurate, and sometimes simply slop, but can it be creative like a human artist? 

We have all seen that AI can produce any kind of media including text, pictures, videos, computer code, and beyond. This is what is so different about the generative AI of today – it creates new things that did not exist before we asked the machine to produce them. The big question is, are these products the results of a creative mind or are they a new form of copying, or even intellectual theft? Furthermore, creativity has been thought to be one of the highest achievements of the human mind. Who can look up at the Sistine Chapel ceiling and not stand agape at the creativity and mastery that came from the mind of Michelangelo? Is there something special about being a great artist, and can that something be captured in code and implemented by a machine? 

Just a few hundred meters from the Sistine Chapel and the many thousands of other masterpieces of art collected in the Vatican Museum  stands the Casina Pio IV, a three-story building set in the beautiful gardens of the Vatican City. This stunning Italian villa was first constructed in 1558 but today houses the Pontifical Academy of Sciences, founded by Pope Pius XI in 1938. There, on September 1-3, 2026, a gathering of philosophers, computer scientists, social scientists, artists, theologians, neuroscientists, and others gathered for the International Center for Consciousness Studies (ICCS) meeting to discuss “Creativity: Minds and Machines” (https://hardproblem.it/). 

This gathering was tasked with diving into the hard problem of machine consciousness and creativity. I was invited to attend the conference due in part to my work with the University of Palermo, where we have been studying the effects that AI and robots have on human moral creativity when it comes to making care decisions for patients suffering from dementia. I presented this work on a panel with my Italian colleagues to open the conference. Personally, I have my doubts that machines available today have the ability to be full moral agents, but I do allow for them to be what philosophers call Ethical Impact Agents, meaning their actions can and do influence the outcomes of ethical decisions made by the human users of these systems. 

The discussions presented by the attendees were wide ranging, but one idea that linked them might be the claim that we have a habit of underestimating the abilities of entities around us while overestimating those same abilities in ourselves, so a kind of intellectual humility is in order. This might suggest that the genius of Michelangelo can be reduced to statistical probabilities playing out on a sufficiently large neural net trained on vast amounts of data, not the divine spark of genius we like to think he was privy to. Personally, I feel that a telling problem was that nowhere in the program did anyone invite an AI or robot artist to speak about their own work and inspirations. Shouldn’t an artist be conscious of their own work and be ready to defend it in the face of critics? At this point in time I believe we can cogently speak of the role that AI can play in human creativity, but it is too soon to suggest that AI is creative in the same way that the best human artists are.

In his recent encyclical “Magnifica Humanitas” (“Magnificent Humanity”), Pope Leo reminds us that, “Our first task is neither to demonize nor idolize technological tools, but to utilize them on the basis of a fundamental principle, namely that truth is a common good and not the property of those with power or influence. We must therefore promote an ecology of communication” (paragraph 137). It is important that communications such as those at the ICCS 2026 in Rome and the Vatican City are happening and that you too can join in this important work here at Sonoma State University.

Note: No AI tools were used in the creation of this article – any slop therein is the product of this human mind

John P. Sullins

Computer Science

Sonoma State University 

Jeff Keating keatingje@sonoma.edu