STAznanost

A year after ChatGPT launch, benefits and pitfalls abound

Ljubljana, 21 December - The development of artificial intelligence (AI) has brought about many advantages but people need to adopt a critical perspective when using it, experts said at a round table hosted by the Slovenian Press Agency (STA) just over one year after the launch of ChatGPT, the most widely known AI tool.

Ljubljana Vasilka Stancin, head of the Institute for International Law and International Relations at the Ljubljana Faculty of Law, speaking at a roundtable on AI development organized by the STA as part of the STAscience project. Photo: Božidar Kolar/STA

Ljubljana
Vasilka Stancin, head of the Institute for International Law and International Relations at the Ljubljana Faculty of Law, speaking at a roundtable on AI development organized by the STA as part of the STAscience project.
Photo: Božidar Kolar/STA

Ljubljana Art director of the Kapelica Gallry art venue in Ljubljana Jurij Krpan speaking at a roundtable on AI development organized by the STA as part of the STAscience project. Photo: Božidar Kolar/STA

Ljubljana
Art director of the Kapelica Gallry art venue in Ljubljana Jurij Krpan speaking at a roundtable on AI development organized by the STA as part of the STAscience project.
Photo: Božidar Kolar/STA

Ljubljana AI expert Marko Grobelnik speaking at a roundtable on AI development organized by the STA as part of the STAscience project. Photo: Božidar Kolar/STA

Ljubljana
AI expert Marko Grobelnik speaking at a roundtable on AI development organized by the STA as part of the STAscience project.
Photo: Božidar Kolar/STA

Ljubljana Vitomir Štruc from the Laboratory for Machine Intelligence at the Ljubljana Faculty of Electrical Engineering speaking at a roundtable on AI development organized by the STA as part of the STAscience project. Photo: Božidar Kolar/STA

Ljubljana
Vitomir Štruc from the Laboratory for Machine Intelligence at the Ljubljana Faculty of Electrical Engineering speaking at a roundtable on AI development organized by the STA as part of the STAscience project.
Photo: Božidar Kolar/STA

Ljubljana Olga Markič, philosopher and professor at the Ljubljana Faculty, speaking at a roundtable on AI development organized by the STA as part of the STAscience project. Photo: Božidar Kolar/STA

Ljubljana
Olga Markič, philosopher and professor at the Ljubljana Faculty, speaking at a roundtable on AI development organized by the STA as part of the STAscience project.
Photo: Božidar Kolar/STA

It is vital that people, especially the young generation, are taught about how AI tools work so that they know how to benefit from such technology and what its dangers are, said philosopher and professor at the Ljubljana Faculty of Arts Olga Markič.

AI researcher Marko Grobelnik said there had been a huge leap in AI development within the last year.

"The things we can do today could not have been done a month ago, and what we could do a month ago could not have been done tree months ago," he explained, adding AI is much more than just OpenAI's ChatGPT platform.

Various corporations are increasingly staking on the AI models' multimodality, said Vitomir Štruc of the Laboratory for Machine Intelligence at the Ljubljana Faculty of Electrical Engineering.

He expects for all types of modalities, including speech, sound, language, and images to be progressively integrated into a coherent system in the future.

Different AI models are also increasingly applied to arts, but many artists have already succumbed too much to certain tools, such as DALL-E and Midjourney, said art director of the Kapelica Gallery Jurij Krpan.

Such artworks lack creativity, depth and all seem alike, he illustrated. However, artists are still most interested in models and modules, that is how to adapt a tool and use it to create something.

Turning to copyright law, Vasilka Sancin, head of the Institute for International Law and International Relations at the Ljubljana Faculty of Law, said that people are always copyright owners and AI is always used merely as a tool.

She highlighted the pitfalls of using AI in a legal setting, noting that as AI technologies are becoming more widespread, it could happen that a person requests the court to make a decision without using such technologies.

"Therefore in law some are already advocating for a new human right to be recognized, that is the right to human decision-making," she said.

Štruc said that while regulation can restrict the development of AI it also offers many opportunities, such as the development of new tools that ensure compliance with the law.

He highlighted the right to be forgotten, which gives individuals the right to request that access to their personal or company data is restricted for the general public.

"What we are trying to do here is develop algorithms that remove from the model ... a certain piece of information. This is a very young domain that has the potential to develop new products," he said.