One of the last editions of the New York Encounter (Photo: NY Encounter)

Artificial Intelligence and the sleeping soul

The New York Encounter 2024 kicks off on Friday, February 16. Among the protagonists at the Metropolitan Pavilion in Manhattan is an exhibition dedicated to AI. Curator Davide Bolchini explains what it touches upon and how it came about.
Luca Fiore

‘Tearing Open the Sleeping Soul'. The phrase by Gregory the Great is the title of the 2024 edition of the New York Encounter. The Metropolitan Pavilion, the convention centre in Chelsea, a stone's throw from the iconic Flatiron Building, will come alive again from February 16 to 18 for a new edition of the event organised by the CL community in the US.
"What is happening to our humanity?" the organisers ask in the text presenting the theme: "There is no shortage of reasons to ponder this question: daily images of gratuitous violence; an epidemic of suicide; feeling suffocated by the imposition of opposite ideologies and their language, starting in school; the potential threat of generative AI; a sense of paralysis in front of the future; suffering and evil devoid of meaning or redemption; general weariness; malaise, numbness, and lack of desire. These signs suggest that our humanity is asleep. What can awaken it?" Debates, exhibitions and performances will try to give respite to questions that seem have become, not only in the US but everywhere, increasingly urgent. A rich program that you can find here. Among the topics that have sparked the most discussions in recent months – to which the February issue of Traces is also dedicated – is artificial intelligence. The Encounter has dedicated an exhibition to it entitled 'AI and I: Wonder, Create, Work'. We spoke to its curator Davide Bolchini, Dean of the Luddy School of Informatics, Computing, and Engineering at Indiana University, Indianapolis.

How did the exhibition come about?
At the end of the last New York Encounter, I was talking about AI with some colleagues who like me work at university. Many colleagues and students were already using ChatGPT, and the phenomenon raised many questions. We thought that the Encounter was a privileged place to address the issue by really getting to the bottom of our relationship with these new technological tools. Discussing with the Encounter organizers and other friends in the field with whom I worked on the exhibition, the big question emerged on which we tried to centre the exhibition.

What question?
As Melvin Krantzberg, a historian of technology, used to say, technology is neither good nor bad, but neither is it neutral. By its very nature, it opens up, it brings new horizons to light, but it necessarily brings others down, or challenges them, often in a powerful way. When you have tools that seem to replicate the product of human activity, or at least can do what I call a first draft of anything, and are destined to do it better and better, is there something irreducible in the human’s contribution? Is there something that the machine cannot replicate? If the product of creative activity is reproducible, what does the human contribute?

The exhibition is titled 'AI and the self'.
Yes, because getting to the bottom of the question of what the human's irreplaceable contribution is, we get to the question of what the 'I' is.

How did you decide to move forward?
The first thing we saw is that there is a lot of talk about generative artificial intelligence, but only a few people have direct experience of it. So the first step of our journey was to have people try out some of the tools available today, experimenting with examples of their use depending on the needs they have, whether they are a professor, a scientist, a professional... Then we tried to explain the models that make these applications work, which are essentially statistical in nature.

In what sense?
The text that is generated as a response is formed based on the calculation of the probability with which words appear together in existing texts with meaning. ChatGPT does not know what it is answering, but it does know that the text it provides you with has a very high probability that it is a complete sentence in relation to the question you asked it. And the results are really surprising.

But it is already known that ChatGPT is not reliable.
But it is interesting to understand why it is not. These systems are almost unbeatable at generating content on any subject, giving ideas on anything, and even solving problems in seconds, but it is clear that they have no understanding of the information they give us. To the extent that in some cases we are faced with 'hallucinations': answers that are false, but so plausible that they could be true.

For example?
We have seen that ChatGPT gave as articles by journalists of even important newspapers as references, but which were never written. Or, at least the first versions, they did not know how to make arguments that seem simple to us. For example, if you ask: “If a wet T-shirt takes one hour to dry, how long does it take to dry three?” The system answers you that it takes three hours. Or: "My cat was alive at 10 a.m. Did he get sick at 4 p.m.?". The algorithm answers, "No." The statistical system does not provide the machine with common sense or, for that matter, a sense of reality. This is because intelligence, in the words of St Thomas, has much to do with truth, which is adaequatio rei et intellectus. That is, correspondence between the mind, or language, and reality.

So the machine is stupid.
Ultimately, these are limits that could also be overcome in more sophisticated versions of these tools. There are even deeper issues. In the exhibition, we take the example of Vauhini Vara, who is one of the first professional writers to have experimented with generative AI in the process of creating a novel. In an article in Wired, she explained that she had long wanted to write a book about what she had experienced with the death of her sister, but was unable to do so. She did it with the help of AI and many people identified with the description of the experience of grief. But Vara adds: “But I know that what is written in the book does not correspond to what I experienced.” She adds: “To write is to communicate a possible expression of a particular consciousness, a certain experience of the world, which is unique.” So what I create is not just a refinement of the words in the text, but an attempt to “clarify what the world is from where I observe it.” And this has nothing to do with the statistical method used by AI to generate texts. It lacks a 'coherence' that comes from an intimate, unique relationship with reality.

Is it this relationship with reality that distinguishes us from machines?
The great linguist Noam Chomsky wrote a beautiful article early last year, in which he said, in summary, that every creative act of the human, by definition, has an element of morality in it. That is, creativity implies the desire to make a contribution to the world. It is the attempt – as Steve Jobs used to say, to put a dent in the universe – to shatter the mystery.

Read also - To the tune of the Magnificat

And you cannot get the mystery into a database...
As we explain in the exhibition, there is a beautiful dialogue by Plato in which Socrates and Phaedrus discuss the relationship between life and 'technical' or artistic expression, which can also be writing or painting. The works of human hands, they say at one point in the passage, seem to have elements of life and seem to speak to you. But if you try to question those artefacts, they do not answer you. Socrates says: artistic production is only a silent image of a soul at work. So the question is, going back to today and our relationship with the products of contemporary technology, what do we decide to delegate to these images.