Researchers have developed a method to train language models using messenger RNA (mRNA) sequences from over 25 species. This approach allows for the creation of large-scale, high-performance language models at a significantly lower cost than traditional methods. The mRNA-based models are trained on a dataset of 1.5 billion parameters and demonstrate state-of-the-art performance in various natural language processing tasks. --- Why it matters: This breakthrough matters to AI researchers because it provides a more affordable way to train large-scale language models, which can be used for a wide range of applications, from chatbots to text analysis tools. Source: https://huggingface.co/blog/OpenMed/training-mrna-models-25-species