
Google has now shared more information about the Universal Speech Model (USM). The USM is a system that has been described as a "critical first step" in realizing the company's goals, which is now moving closer toward its goal of building an artificial intelligence (AI) language model that supports 1,000 different languages to beat ChatGPT
Google in November last year had announced its plans to create a language model supporting 1,000 of the world's most-spoken languages while also revealing its USM model.
USM has been explained as a family of state-of-the-art speech models with 2 billion parameters trained on 12 million hours of speech and 28 billion sentences of text, spanning 300+ languages.
"USM, which is for use in YouTube (e.g., for closed captions), can perform automatic speech recognition (ASR) not only on widely spoken languages like English and Mandarin, but also on under-resourced languages like Amharic, Cebuano, Assamese, and Azerbaijani to name a few," Google said in a blog post.
Currently, the tech giant claims that USM supports over 100 languages and will serve as the "foundation" for a much larger system.
Meanwhile, in a bid to compete with the popularity of ChatGPT, Google rushed in to launch Bard. The response for Bard has been underwhelming. Employees of the company have described the rollout as 'botched'.
Google has said that it will also include Bard's underlying technology LaMDA in Google Search. It will essentially let users search for queries and get results based on SEO optimization. Additionally, it will also have a separate chatbox to include conversational results based on web results.
However, Sundar Pichai has been on a mission to make it work. According to a CNBC report, Pichai in a company-wide mail asked the workers to test Bard and spend some time with it.
Read: 'Google was hesitant': Alphabet chairman opens up about ChatGPT-rival Bard
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine
Copyright©2025 Living Media India Limited. For reprint rights: Syndications Today