Artists question the place of artificial intelligence in visual art

Reading time: 4 minutes

According to Investment Ontario, the province will become “a true artificial intelligence (AI) hub, not just in Canada, but on a global scale”. If the reputation of AI has grown over the years, it has also raised some questions, especially in the artistic world in French Ontario, where there is no escape from it. Dehumanization of art, devaluation of manual know-how, uncertainty about copyright ONFR+ make the point.

“AI refers to several scientific disciplines in computer science, in mathematics, with techniques that have different goals. It can be language, anything visual, etc. explains Céline Casters-Renard, a professor at the University of Ottawa which specializes in AI and the regulation of new technologies.

In other words, it is an algorithm that, in an artistic use, designs a finished work from ingredients dug up on the web.

AI: with or against artists?

“Where some challenges start is that AI can do things and have a certain originality, although these tools are based on already existing data”, believes the expert.

It’s the borrowing of data that causes the problem, according to Franco-Ontarian illustrator and graphic designer Josée Lavoie. “Of course there are always many things that pose a risk in the world of graphic design and art,” he pointed out, giving the Fiverr software as another example, “but mainly for the description that it is dangerous”.

The latter emphasizes his particular discomfort with the potential rights theft that artists face, following the publication of their work online.

Carys J. Craig, attorney and professor of intellectual property law at York University. Photo credit: Joncarlo Lista

In reality, “there’s not much that can be done,” says Dr. Carys J. Craig, an attorney and professor of intellectual property law at York University. According to the latter, it is very difficult to know how the elements of the works are used as well as where they are taught.

“If the artists’ visuals are found, they are usually absorbed into a dataset from which an AI can generate an image,” he said. “So an artist will never know if their work is included or what work is included and actually these datasets often consist of thousands, tens of thousands, hundreds of thousands or even millions- million personal computers. »

Mrs Castets-Renard agreed for her part: “I don’t know whether to worry, but the boundaries or the concepts of copyright are challenged. »

The concern of Héritier Bilaka, a visual artist from Ottawa, lies in the use of AI itself. According to the man who has been painting portraits for 14 years, “if it’s enough to tap a keyboard and then say we’ve created something, I don’t know if the next generation will still have a sense of effort which is linked to the manual. knowledge, physical contact at work”.

Heir Bilaka, visual artist from Ottawa. Respected

The artist from Congo claims to have a different concept of what art “should” be. “For me, art is not something that can only be visual or technical. It is especially related to the soul and to spirituality, because it carries a living energy”, he continued.

However, “for a long time in copyright, we admit that the author (artist) is not necessarily the creator of the material”, corrected Ms. Casters-Renard. In the case of creating AI works, the human always remains the designer who orchestrates a finished result, he explained. This is why some think of AI as a tool, which legitimizes its role in more than one art field.

So is digital artist Karen Vanderborght, who is experimenting with AI creativity software like Midjourney, Dall-E and Dream Studio, all gaining popularity. The latter sees the spontaneous and rational aspect of AI as a phenomenon where it artistically enriches collaboration.

Karen Vanderborght, digital artist.

“I just like how AI gives me strange suggestions. Me, I’m not interested in thinking of it as a process of excellence, I see it more as an artistic game”, reveals one who is interested in more than a new technology , such as Augmented Reality and Virtual Reality.

The latter believes that “we will always have a market where people want a painting made by someone’s hand”.

Céline Castest-Renard shares the same opinion: “It is likely to think that, at the moment, there is no pure and simple exchange of people but certainly a change of tasks. »

A limit to AI creativity

If it is statistically difficult to prove that the rise of AI is harmful to artists, it cannot be denied that it will never completely replace them, according to Ms. Casters-Renard. The reason is completely unanimous among all participants: it cannot include the authenticity of a human approach.

The professor gives the example of a company that wants to work with an artist compared to an AI software: “If everyone has the same graphic design because the tools capture the spirit of the time, everything is mainstream, companies may not be happy with the result. It does not allow them to differentiate themselves or be different from others. »

And to prove that changes in the market are often, but not always counter-constructive for people: “It can only push us a little more with our particularities, our skills and our expertise”, he admits . , referring to his personal experience as a teacher.

Leave a Reply

Your email address will not be published. Required fields are marked *