The installation aims to open the discussion about artificial intelligence through interaction with it

MONTREAL – Talking with an artificial intelligence entity to understand it better: this is the challenge of the installation CHOM5KY vs CHOMSKY, which will be presented from September 6 at Espace ONF, in Montreal.

For 25 to 30 minutes, trial participants will wear a virtual reality headset, immersing them in an immersive, futuristic-looking world. They will be able to ask questions and talk to CHOM5KY, an artificial intelligence entity created in part using the digital traces of American linguist and professor at the Massachusetts Institute of Technology (MIT), Noam Chomsky.

“We can talk to this Chomsky, have a conversation with Chomsky, ask questions, and see how he answers. So far, so good. But this entity also invites us to go through different stages, sometimes alone, sometimes with others, and sometimes in cooperative games, to better understand what distinguishes our human intelligence from machine intelligence,” explained Sandra in an interview. The Chomsky Experience, who taught at MIT For several years.

His new project aims to demystify the work of artificial intelligence, in a period when this technology occupies an increasing amount of space and thus raises concerns.

“It’s still very easy to see how (AI) works, and then understand that it’s not magic. There’s no intention, there’s no intelligence behind it, it’s really just a gifted way of formulating sentences that seems to us to meet a need,” Ms. Rodriguez explains. When we understand how it works, we are not deceived, nor are we imitating anyone’s intelligence.

The VR experience therefore allows us to look “under the hood” of artificial intelligence, in order to understand how responses are produced. CHOM5KY shows people who interact with him what other formulas he could have used to answer a question, or what synonyms he could have used, for example.

After the virtual reality portion of the experience, visitors will be invited to learn more about artificial intelligence in a physical environment.

“CHOM5KY, in the process, (invites us) to discover how machines work. This project also leads us to understand ourselves better,” says Ms. Rodriguez, who has been working on this project since 2016.

“One of the world’s most digital thinkers”

Why did you choose Noam Chomsky to implement this project? The idea came about when Ms. Rodriguez was a professor at MIT, and a young researcher told her about his research project.

She says: “He told me: There are a lot of digital influences on Noam Chomsky, and I am sure that if I can understand and imitate the way he speaks, I will be able to understand how his brain works.”

Then “the irony of this proposal hit her hard.” “The great irony,” she continues, “is that Chomsky’s theories say exactly the opposite.” According to Mr. Chomsky, it is impossible to imitate human intelligence, because we still know very little about it, and therefore we cannot imitate something we do not know.

For the professor, it would have been better to use the numerous traces of “one of the world’s most digitized intellectuals”, including conferences, interviews and some of his courses, to create an artificial intelligence entity that shows its flaws.

“Within these monuments is his legacy: his belief that human creativity is what creates our freedom. It suddenly becomes a perfect way to create an AI entity that reminds us that it is limited, reminds us of its potential pitfalls, and reminds us of Chomsky’s message about our ability to question, what makes us unique as humans, and creativity, (and) we must not lose it,” Ms. Rodriguez explains.

Noam Chomsky’s digital artifacts are available online, royalty-free. However, Ms. Rodriguez informed Mr. Chomsky, now 94, about her project, and he decided not to participate in it.

More transparency is needed

During a conference on Wednesday, as part of the imminent launch of CHOM5KY vs CHOMSKY, Sasha Lucioni, researcher in ethical and sustainable AI for HuggingFace, highlighted the importance of companies developing AI entities showing more transparency.

“Companies are talking less and less about how they collect their data, how many people they trained their algorithms, and (how much) they got paid. (…) Unfortunately, I find that instead of calling for transparency, which will eventually lead to regulation (it’s Secret),” says Ms. Lucioni.

It reminds us that artificial intelligence is not created by “magic,” but on the contrary, it involves a lot of human intelligence. She points out that programs like ChatGPT rely on data that humans post online (such as news sites), and many people are hired to improve the program.

Ms. Lucioni emphasizes that the deployment of AI models will be useful for researchers, but also for the population to understand how this technology works.

The CHOM5KY and CHOMSKY experience will be offered until October 15.