Metaverse is a rapidly gaining a lot of traction, and Meta (previously Facebook) is leading the charge, offering different technology advancements to assist the internet’s future. Meta is working on a pair of unique metaverse gloves that allow users to touch and feel virtual items, as we just learnt. The company has now announced that it is developing an AI-based universal language translation system to help metaverse users interact more effectively.
Metaverse AI Universal Language Translation System
Meta CEO, Mark Zuckerberg presented a variety of new AI-based technologies that his business is currently building for the metaverse in a recent Inside the Lab: Building for the Metaverse with AI live webcast. Among other things, Zuckerberg stated that Meta’s goal is to create a universal language translation system that will allow people to connect with others in the metaverse without fear of language barriers.
Zuckerberg added, “The key goal here is to construct a universal model that can combine knowledge across all modalities.” “This will allow for a massive scale of predictions, judgments, and creation, as well as entirely new architectures, training methods, an algorithms that can learn from a huge and diverse variety of different inputs,” he continued.
When it comes to the company’s plans for creating such a model, it claims it is currently working on two independent AI language translation models. The first, “No Language Left Behind,” will be able to learn any language, even if there is just a small amount of text to work with.
“We’re building a single model that can translate hundreds of languages with cutting-edge results and most language pairs — everything from Austrian to Ugandan to Urdu,” Zuckerberg explained.
The company’s M2M-100 model, which was announced in 2020, would be used to construct an AI Babelfish in the second model. “The goal here is instantaneous speech-to-speech translation across all languages, including those that are predominantly spoken; the capacity to communicate with anyone in any language,” Zuckerberg said further adding to his statements.
These are some important claims that Meta is making right now. Despite the fact that the Facebook AI Research team is always working on a variety of AI technologies, the firm continues to face data shortage difficulties.
“Text translation machine translation (MT) systems often learn from millions of sentences of annotated data. As a result, only a few languages dominate the web, and MT systems capable of high-quality translations have been built for them,” the Meta FAIR team noted in a recent blog post.
