Facebook is developing the most powerful artificial intelligence supercomputer named the AI Research SuperCluster (RSC), said Meta. A glimpse of it can be possible by mid-2022 and is being viewed as a major step towards making AI capabilities more powerful.
The project is partly focused on the metaverse initiative of the company and its AI-products and applications are to play important roles, added further the Facebook owner.
As of now, the RSC is capable of creating models for natural-language processing and computer vision. It is believed the same can be used in speech recognition as well as more advanced technologies in the future.
It is being simultaneously hoped that entirely new AI systems could be developed equipped with abilities to provide large group voice translations in real-time even though the members speak in different languages.
A host of technology will be put into use and many are not yet completely realized. The technology can analyze large amounts of data even in concerts, large gatherings and other such loud scenarios. It can understand the voices of people speaking in various languages, accents and dialects as well simultaneously.
Meta has been working on advanced AI technology for nearly a decade and initially started with the Facebook AI research lab. However, the RSC is the most spectacular and largest one. Meta has been working on advanced AI technology for nearly a decade and initially started with the Facebook AI research lab. However, the RSC is the most spectacular and largest one.
However, the actual plan kicked off in early 2020 to build a new computing infrastructure and train models to analyze data sets of even the size of exabyte, which is equivalent to 36,000 years of videos.
Meta confirmed the RSC is in function now and to be only growing further in the future to increase the AI training performances by more than 2x. This will help in the improvement of the AI as well as its existing services.
The RSC uses open source as well as other publicly available data sets, ensuring Facebook that the research is effective in practice and includes real-world examples.
Metaverse is a 3D virtual world equipped with a network and basically focused on social connection. The name was first coined in 1992 with a science fiction novel named Snow Crash. Several components of it are being used in internet-enabled video games. The access points required are augmented reality (AR), virtual reality (VR), mixed reality and virtual world technologies. Information privacy is of concern here as the personal information of users will be collected by companies. Additionally, virtual crime may be a big safety issue.