A team from the UPV presents a study on ethical reasoning of AI at a conference in the United States

VALENCIA, 2 Mar. (EUROPA PRESS) -

A team from the GTI-IA research group of the Valencian University Institute for Research in Artificial Intelligence (VRAIN) of the Universitat Politècnica de València (UPV) has presented a work on ethical reasoning of artificial intelligence at the annual conference organized by the Association for the US Advancement of Artificial Intelligence (AAAI), one of the most prestigious international conferences in this field.

The article, under the title 'Moral uncertainty and the problem of fanaticism', has been prepared by VRAIN researchers from the UPV, José Such and Natalia Criado, and collaborators from King's College London (United Kingdom), Jazon Szabo and Sanjay Modgil. And it develops how, through a mathematical formalism, artificial intelligence can be made to reason about different points of view and ethical theories, the university indicated in a statement.

"We are telling the AI ​​that when deciding, take into account and add various ethical aspects, not just an ethical perspective. And in this way, its final decision will be more aligned with what a greater number of people consider ethical , and will move away from none of the ethical theories disproportionately dominating over the others, which would lead AI to more fanatical behavior," explained José Such.

Along these lines, he added that a question is "very important in order to develop artificial intelligence systems that are capable of aligning with what people consider to be ethical behavior."

This white paper is based on research from the VAE (Value-awareness engineering) project. Its objective is, precisely, to develop tools and technologies to build intelligent systems that understand human values ​​and guarantee that their behavior is aligned with those values.

The project is financed by the Ministry of Science, Innovation and Universities and the EU through Next Generation Funds and has a duration of two years. Its main researcher is Natalia Criado, in the team formed by Vicent Botti and José Such from the GTI-IA group of VRAIN of the UPV.

The AAAI conference on Artificial Intelligence, which was held from February 20 to 27 in Vancouver (Canada), has the objective of promoting scientific exchange between professionals, researchers, scientists and students in this field and its disciplines and brings together the most brilliant in AI throughout the year.

This is one of the conferences with the longest history in AI, also organized by one of the associations with the longest history in this technology. The articles presented in this forum are very limited due to the demands of high scientific standards.

The Valencian Research Institute for Artificial Intelligence (VRAIN) of the UPV is made up of eight research groups that have more than 30 years of experience in different lines of AI research.

The creation process of VRAIN began in 2019, the result of the union of six research groups. In 2020, it merged with the PROS Software Production Methods Research Center and in 2021 it was finally established as a University Research Institute with the approval of the Generalitat Valenciana.

Currently, it has 178 researchers divided into nine research areas. These nine areas on which its research activity revolves mean that its developments are applied to a large number of strategic sectors such as health, mobility, earth sciences, smart cities, education, social networks, agriculture, industry, privacy/security, robots self-employed, services and energy, and environmental sustainability among others.

These activities have been financed by more than 135 projects obtained through competitive financing, mainly from the European Union, but also from the National Research Plan, the Valencian Research Plan and Technology Transfer Projects.