Google CEO Sundar Pichai has announced the launch of a new AI language model called “BART” (Bidirectional Encoder Representations from Transformers) that aims to take on OpenAI’s GPT series of language models, including ChatGPT. The BART model is designed to better understand and generate natural language and is said to be capable of a wide range of language tasks, such as text summarization, translation, and question-answering.
According to Pichai, BART is different from other language models because it can understand language in a more “holistic” way, which allows it to better capture the meaning and context of words and phrases. This is achieved through the use of a new neural architecture that combines the strengths of both transformer and autoencoder models.
Pichai also mentioned that BART has been trained on a vast amount of text from across the web, including books, articles, and websites. This training has allowed BART to develop a deep understanding of the nuances and complexities of human language.
Google plans to make BART available to developers and researchers, allowing them to build new applications and tools that can benefit from its language understanding capabilities. The company is also exploring the potential of using BART to improve its existing products, such as search and Google Assistant.
In conclusion, Google’s launch of BART is a significant development in the field of natural language processing, and it will be interesting to see how it compares to existing language models, including ChatGPT. With its unique architecture and extensive training, BART has the potential to be a game-changer in the world of AI and language understanding.
