IBM and Spain are collaborating to build large language models in Spanish
The new agreement between IBM and the government of Spain aims to build the world's leading large language model for Spanish.
The new agreement between IBM and the government of Spain aims to build the world's leading large language model for Spanish.
Generative AI has impacted just about every industry, almost overnight. But much of the work to date has been conducted in English.
Spanish is the second-most spoken language in the world, with nearly half a billion native speakers across the globe. Without robust tools trained on Spanish language materials, hundreds of millions of people, and countless businesses, would struggle to get the same level of access to modern AI applications trained solely in English. The Spanish government and IBM are looking to change that.
In a new effort announced today, the government of Spain and IBM said they would come together to develop and deploy cutting-edge AI technologies. This will include a new open-source large language model (LLM) in Spanish. Together, they plan to revolutionize how Spanish-speaking countries and organizations can do business in the age of AI.
Building upon IBM’s AI tools and open-source prowess, the goal is to enable Spanish-speaking communities to contribute to strengthening the proposed model for additional use cases. The collaboration will also work to build other Spanish-language models, including smaller models for related dialects like Catalan, Basque, and Galician.
“We will leverage IBM’s deep investments across the entire AI stack, from compute to models, we will enable an open community to drive the full potential of AI technologies,” said IBM’s Darío Gil, senior vice president and director of IBM Research. “Together, we can ensure that AI is used ethically to drive sustainable growth in Spain, and all Spanish speaking countries worldwide.”
