Odaily Planet Daily News According to people familiar with the matter, Baidu (BIDU.O) plans to release the next generation of artificial intelligence models in the second half of this year. Ernie 5.0, known as the "base model", will have major enhancements in multimodal capabilities, the source said, but did not specify its functions.
"Multimodal" artificial intelligence can process text, video, images and audio and combine them together, and can also convert across categories - for example, text to video and vice versa. The base model can understand language and perform a wide range of tasks, including generating text and images, and communicating in natural language. Currently, Chinese companies are competing to develop innovative artificial intelligence models to compete with OpenAI and other American companies. (Jinshi)