In a surprising move, the Spanish government has announced an agreement with IBM to create an AI model in Spanish. This collaboration is part of the government’s efforts to develop alternatives to large Anglo-Saxon models and sparked debate within the AI community about the use of public funds for these models.
While some advocated for models developed by established companies like OpenAI or Meta, others supported the creation of native models. The decision to partner with an American technology company like IBM was met with surprise in the Spanish technology sector, particularly as it contrasted with the push for European strategic and technological autonomy seen in countries like France and Germany. These countries have invested in startups within the AI field to enhance their capabilities and drive innovation.
The Spanish government has also been advised to develop a supercomputing strategy optimized for building and serving AI models tailored to its needs and ambitions. This shift in strategy represents a departure from previous efforts to focus on European technology for supercomputing and quantum computing capabilities in Spain. The decision last year to purchase quantum computers from non-European companies had sparked criticism from within the government, but the government defended this choice based on the advanced capabilities of IBM’s technology.
These recent decisions reflect a changing approach to AI strategy under the current administration, which includes developing a National Agency for Supervision of Artificial Intelligence. Overall, the Spanish government is navigating a complex landscape in AI development, balancing innovation and strategic autonomy with considerations around public spending and international partnerships. The agreement with IBM marks a significant step towards developing AI models trained in Spanish and tailored to Spain’s specific needs.
