Spanish government gives 5 million euros so that future AI will ‘speak’ Spanish

The Royal Spanish Academy, which is the official guardian of Castilian, will lead the project that aims to ensure that the language retains a leading role in the digital sphere, too

On Tuesday, The Spanish Council of Ministers approved a direct grant of 5 million euros to the Royal Spanish Academy (RAE) for the implementation of the Spanish Language and Artificial Intelligence (LEIA) project. The initiative has the goal of broadening the importance of Spanish language in the rapidly developing digital world by ensuring that AI technology will be able to quickly and properly process that language.

The project seeks to promote the development of an industry based on language technologies. These are things, such as natural language processing, automatic translation and conversational systems. These tools can allow companies and products in Spanish to compete in the technological revolution that is currently underway.

The Royal Spanish Academy is the formal state institution in Spain that safeguards the correct usage and rules of the language.

The 3^rd most used language in the world, but are machines fluent in it?

The collaboration between the Government and the RAE focuses on taking into consideration the language - also known as Castilian - which is the third most used language on the Internet after English and Chinese. It is a fundamental element for the development of chatbots, virtual assistants, simultaneous translation and reading comprehension tools. 

Through this action of the Government and the RAE, linguistic resources, tools and models will be generated that will be made available to research centres, companies and users in an open, accessible manner and in suitable formats for use and reuse. This ‘learning’ process will represent an important knowledge transfer to the economy.

Among the many planned activities, the creation of an observatory of neologisms and technicalities stands out where, through AI, these types of words are automatically collected in real-time. That means AI will be able to understand how youngsters speak on the street rather than quote Cervantes.

In addition, linguistic content verification and automatic response tools will be created and developed. For this, it is necessary to properly organize and classify all the data and create tools that allow it to be exploited optimally. 

The results obtained thanks to the execution of this collaboration will benefit the entire Spanish-speaking community, made up of 591 million people, 7.5% of the world's population.