Latam GPT: el modelo de inteligencia artificial que busca representar a Latinoamérica Desarrollado en Chile y respaldado por otros países, este modelo de IA promete ofrecer una visión más precisa de la historia y la cultura de la región.
Latin America is about to take a significant step in the development of artificial intelligence (AI) with the creation of Latam GPT , a language model designed to represent the history, culture and linguistic diversity of the region.
According to CNN , the project is promoted by the Ministry of Science, Technology, Knowledge and Innovation (CTCI) of Chile and the National Center for Artificial Intelligence (Cenia), with the support of experts and institutions from various countries, including Mexico, Argentina, Colombia, Ecuador, the United States, Spain, Peru and Uruguay.
One of the main problems with current AI models is that, although they possess vast knowledge, they do not always accurately understand the Latin American context. Phrases, idioms and cultural references are often misinterpreted or generalized.
This model has been developed with 55 trillion parameters and a database of three trillion tokens, a capacity similar to that of OpenAI's ChatGPT 3.5.
The difference lies in its focus on Latin American content, which will allow it to respond more accurately to historical, social and cultural issues in the region.
"LLMs (Large Scale Language Models) are rapidly becoming part of our daily lives. But what happens when we ask questions about Latin America? Although current models are of high quality, their understanding of the Latin American context can be improved. That is why we believe it is imperative to develop a regional model that accurately reflects our history and culture," the project's official website mentions.
Unlike other AI models that belong to large technology corporations, Latam GPT is conceived as a public and open system, which will allow developers and researchers to customize applications based on its architecture.
For his part, Rodrigo Durán, general manager of Cenia, explained that the main objective of the model is to provide a powerful tool to address specific problems in Latin America and the Caribbean, such as data analysis in public policies, education, environment and economy.
Related: Almost half of workers in Latin America use AI
Where did the funding for Latam GPT come from?
The development of Latam GPT has been made possible thanks to an investment of $550,000, of which the Andean Development Corporation contributed almost half. In addition, the University of Tarapacá has allocated $5 million for the installation of a supercomputer, accelerating the model training process.
In Mexico, institutions such as the Center for Research in Geospatial Information Sciences, the National Institute of Astrophysics, Optics and Electronics, and the Center for Research and Innovation in Information and Communication Technologies have contributed data and technical advice to improve the performance of the model.
Latam GPT is still in the training phase, but its team has opened up the possibility for the public to participate in this process. Through the project website, anyone interested can request access and contribute by sending texts in Spanish that reflect the linguistic nuances of each country.
