Skip to main content

Artificial Intelligence is not a digital Aristotle, but it can help you find material

Instead of artificial intelligence, we could use the term virtual intelligence, as powerful machines are still far from being intelligent, and instead of an ethics of artificial intelligence, we could talk about the ethics of those who use modern technologies, said Igor Papič, Minister of Higher Education, Science and Innovation, in his opening remarks on the second day of the Global Forum on the Ethics of Artificial Intelligence.

He said that artificial intelligence is not a threat but a tool that can be used to our advantage. However, as modern technologies have a strong impact on our communities, engineers and sociologists need to work together, according to Minister Papič.

Mariagrazia Sqicciarini from UNESCO, moderator of the first panel on Inclusive AI for solutions that improve the quality of life for all, asked the gathered how the international community can address the pressing issue of the unequal distribution of resources and the large disparities in earnings that private companies make from data collection.

Renata Vicentini Mielli, from the Brazilian Ministry of Science, Technology and Innovation, said that while the problems associated with technological inequality were not new, large digital platforms had only exacerbated them and accelerated the trend of increasing the concentration of resources in the hands of a minority. Investing in data centres is key, she said, to ensure that countries maintain sovereignty over data, particularly in areas such as health, education, research, and security. Data sovereignty is important for the state, she said, because it is a way for states to effectively limit the concentration of power, money, and wealth in the private sphere.

Inma Martinez from the Global Partnership on Artificial Intelligence (GPAI) recalled that for more than twenty years, private companies have been collecting data in the digital world without regulation. We cannot repeat the past, but we can make sure that we better preserve essential social values in the future, she said. "There is no need for a private company to collect data from your state-issued passport," Martinez believes.

The design of the development and utilization of AI must be democratic, not just confined to the domain of a small number of private companies. However, as universities cannot compete with private companies, countries should ensure with investments that universities and research institutions can also create and develop AI, Martinez suggested.

When we talk about artificial intelligence, we are responding to something we know very little about, says Matjaž Gruden from the Council of Europe, where the Convention on Artificial Intelligence is being developed. Gruden suggests that to answer complex contemporary questions, we should look to previous, proven, and trusted models that are worth transferring and applying. For example, he believes that the principles of the market, of commodity exchanges, and of limiting monopolistic practices, which have long been known, can also be applied to contemporary phenomena. However, this cannot be done by states alone, and citizens need to be constantly educated about new phenomena.

Gruden said that the European Year 2025 will be dedicated to education, and Slovenia will host one of the four conferences that will be part of next year's events.

Zeng Yi, a member of the Chinese Academy of Science, pointed to the fact that AI is practically non-existent in reducing poverty, eliminating hunger, ensuring access and protecting water, because it does not bring in big profits, even though these are areas that are vital for humanity and would need the help of AI the most.

On data, he suggested that each country should develop its own generative AI based on its own data and in its own language. This will ensure that countries learn about AI in their own language and encourage interactions between cultures. He compared the model to UNESCO's principles of heritage protection: each culture is unique, but they all form a world cultural heritage.

Anyone who wants to learn about traditional cultures cannot talk about it with a chatbot, Yi believes. Artificial intelligence is not a digital Aristotle, but it can help find a lot of material about Aristotle.

Today's second Global Forum debate was based on a bold question: do we need to re-engineer the system?

Huub Jansen from the Dutch Authority for Digital Infrastructure said that the development of AI is indeed happening at lightning speed, "but if you think great things are already happening, forget it–this is just the beginning". We know that we need to control this process because we all want a fair distribution of resources and fair and trustworthy AI, but no one knows how to approach this, says Jansen.

But we cannot wait for something to happen, we need to be proactive. In the Netherlands, a working group has been set up for this purpose, involving many competent authorities. "We don't know exactly where we are going. We just know that we are facing change and that we need to act quickly," says Jansen. They are mainly learning from practice by simply getting on with the work.

Samo Zorc, the coordinator of the preparation of the National Programme for Artificial Intelligence (NpUI), compared the rapid changes and the uncharted territory of artificial intelligence with the situation thirty years ago and the state of computer software at the time of the advent of the web, which greatly accelerated change and required agile thinking. In the transition from traditional policy-making and the need for rapid action, Zorc proposes agile policy-making and agile processes also at the level of organisations and systems. However, as no one can be completely independent today, cooperation and internationalisation, taking into account cultural differences, are, in his view, essential in development issues.