Google is one step closer to creating a 1,000-language artificial intelligence model

Google is developing all sorts of AI technologies, including a universal speech model, which is part of an attempt to create a model that can understand the world's 1,000 most common languages.

While Microsoft and Google argue over whose AI chatbot is better, that's not the only application of machine learning and language models. Along with rumored plans to showcase more than 20 AI-powered products during this year's annual I/O event, Google is moving toward its goal of creating an AI language model that supports 1,000 different languages. In an update on Monday, Google shared more information about the Universal Speech Model (USM), a system that Google calls a "critical first step" in achieving its goals.

Last November, the company announced its plans to build a language model that supports 1,000 of the world's most widely spoken languages, and simultaneously unveiled the USM model. Google describes USM 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 in more than 300 languages.

USM, which YouTube already uses to create closed captions, also supports automatic speech recognition (ASR). It automatically detects and translates languages including English, Mandarin, Amharic, Cebuano, Assamese, and more.

Currently, Google says USM supports more than 100 languages and will serve as a "foundation" for an even more extensive system. Meta is working on a similar translation tool with artificial intelligence, which is still in its early stages. You can read more about USM and how it works in a research paper that Google has posted here.

One application for this technology could be augmented reality glasses like the ones Google demonstrated last year at the I/O event. They would be able to recognize and provide real-time translations appearing right in front of your eyes. However, this technology still seems a bit remote, and Google's misrepresentation of Arabic during I/O proves just how easy it is to get it wrong.