Humanizing AI with language technology: the 'HumanAIze' project kicks off
The initial meeting of the project, held in Donostia-San Sebastián, marks the beginning of a national initiative in artificial intelligence aimed at developing more reliable, ethical, and human-centered language models.
Advancing towards a new generation of Large Language Models (LLMs) that are more human-centered, reliable, and aligned with ethical values. Under this reliable-by-design approach, the HumanAIze project was born, a new initiative with the participation of CiTIUS which held its official kickoff meeting last week in Donostia-San Sebastián.
The meeting took place last Friday at HiTZ Zentroa, the research center of the University of the Basque Country and coordinating entity of the project. CiTIUS researchers Marcos Garcia and David E. Losada took part in the event, which brought together teams from the six institutions that make up the consortium. In this context, David Losada emphasized that “the awarding of HumanAIze in a call with such a low success rate strengthens CiTIUS’s position in flagship projects and highlights the relevance of its lines of research in language technologies.”
The project has been awarded in a call with a very low success rate, which strengthens CiTIUS’s position at the forefront of AI
This first meeting made it possible to align the project’s scientific vision and define the lines of work for the coming months, in a collaborative context that brings together the University of Santiago de Compostela (through CiTIUS and the Instituto da Lingua Galega), along with UPV/EHU, the University of Alicante, the University of Jaén, UNED, and the Barcelona Supercomputing Center.
HumanAIze addresses some of the main challenges of current systems, such as biases, the generation of incorrect information, limitations in cultural understanding, or difficulties in multilingual contexts. To this end, the consortium proposes an interdisciplinary approach that combines artificial intelligence, linguistics, and legal studies.
The project will also promote the development of multilingual and multimodal language technologies, with special attention to efficiency, openness, and their application in real-world contexts. Likewise, new evaluation methodologies will be explored, integrating dimensions such as linguistic quality, fairness, and user experience.
Beyond one-off collaboration, HumanAIze is conceived as a virtual laboratory that structures the joint work of teams with complementary profiles, with the aim of generating significant advances in the development of more robust, interpretable, and inclusive AI systems. At CiTIUS (a center co-funded by the European Union through the Galicia Feder 2021–2027 Program), the project involves a research team coordinated by Pablo Gamallo.
Advancing towards a new generation of Large Language Models (LLMs) that are more human-centered, reliable, and aligned with ethical values. Under this reliable-by-design approach, the HumanAIze project was born, a new initiative with the participation of CiTIUS which held its official kickoff meeting last week in Donostia-San Sebastián.
The meeting took place last Friday at HiTZ Zentroa, the research center of the University of the Basque Country and coordinating entity of the project. CiTIUS researchers Marcos Garcia and David E. Losada took part in the event, which brought together teams from the six institutions that make up the consortium. In this context, David Losada emphasized that “the awarding of HumanAIze in a call with such a low success rate strengthens CiTIUS’s position in flagship projects and highlights the relevance of its lines of research in language technologies.”
The project has been awarded in a call with a very low success rate, which strengthens CiTIUS’s position at the forefront of AI
This first meeting made it possible to align the project’s scientific vision and define the lines of work for the coming months, in a collaborative context that brings together the University of Santiago de Compostela (through CiTIUS and the Instituto da Lingua Galega), along with UPV/EHU, the University of Alicante, the University of Jaén, UNED, and the Barcelona Supercomputing Center.
HumanAIze addresses some of the main challenges of current systems, such as biases, the generation of incorrect information, limitations in cultural understanding, or difficulties in multilingual contexts. To this end, the consortium proposes an interdisciplinary approach that combines artificial intelligence, linguistics, and legal studies.
The project will also promote the development of multilingual and multimodal language technologies, with special attention to efficiency, openness, and their application in real-world contexts. Likewise, new evaluation methodologies will be explored, integrating dimensions such as linguistic quality, fairness, and user experience.
Beyond one-off collaboration, HumanAIze is conceived as a virtual laboratory that structures the joint work of teams with complementary profiles, with the aim of generating significant advances in the development of more robust, interpretable, and inclusive AI systems. At CiTIUS (a center co-funded by the European Union through the Galicia Feder 2021–2027 Program), the project involves a research team coordinated by Pablo Gamallo.