Recent developments in AI, and especially Large Language Models (LLMs) are starting to have a significant effect in our daily lives, and they are bound to have a profound impact on our society, for better or worse. Due to the enormous costs of creating powerful AI models only a few players are leading the field, mostly large American tech companies, leading to a risk of significant concentration of power. The Barcelona Supercomputing Center is one of the few public research institutions with the computational and technical capacity to provide a counterweight to this. In this talk I will present the challenges we encountered and the work being done at the BSC to make open and transparent AI models available to the public, and to ensure that European languages and culture are adequately represented in those models. Leveraging the BSC's capacity to train large and powerful models, we especially focus on producing models, data, and tools that can be used and adapted by anyone, even on a limited budget.
Session 🗣 Introductory and overview ⭐ Track: AI, ML, Bigdata, Python
AI
LLMs
Large Language Models
NLP
Language Technology
Catalan
Spanish