Mark Zuckerberg, CEO of Facebook's parent company Meta, has announced that the company is building what it says will be "the fastest AI supercomputer in the world when it’s fully built out in mid-2022."
"Meta has developed what we believe is the world's fastest AI supercomputer. We're calling it RSC for AI Research SuperCluster," he said in a Facebook post.
Meta, which owns Instagram, Facebook, and Whatsapp, has been involved in multiple controversies regarding privacy violations, and the spread of disinformation through echo-chambers.
A Supercomputer to Power the Metaverse
Meta claims the AI supercomputer, called AI Research SuperCluster (RSC), is already the fifth fastest in the world. It is built from thousands of interconnected processing units.
RSC will help Meta’s AI researchers build new and better AI models to analyze text, images, and video together, as well as develop new augmented reality tools, it says.
Meta also hopes RSC will help it build AI systems that can power real-time voice translations to large groups of people around the world to ensure seamless collaboration.
The social media giant has also claimed that the supercomputer would help identify harmful content on its platforms.
“The experiences we’re building for the metaverse require enormous compute power (quintillions of operations/second!) and RSC will enable new AI models that can learn from trillions of examples, understand hundreds of languages, and more,” Zuckerberg said in his post.
Meta owns Facebook, Instagram and WhatsApp, and boasts nearly 3 billion daily users whose data can be used to train AI models.
Meta recently rebranded itself to signify its push towards building a 'metaverse', an iteration of the internet which leans heavily on virtual reality.
Concerns Over Privacy and Disinformation
Facebook has received heavy criticism for its use of user data. Its AI algorithms reportedly tend to track what users like and funnel them towards similar posts, creating echo chambers and often fueling disinformation and hate speech.
It has also seen numerous privacy breaches, dating back to 2006.
In 2018, Meta (then Facebook) was fined $5 billion and issued a public apology after Cambridge Analytica took user data from the platform to build a system to profile individual US voters and target them with personalised political advertisements.
In 2020, Facebook paid another $550 million to resolve a lawsuit which claimed it collected user biometric data without consent.
"Nothing good can come from all of that computer power in the hands of such a tech superpower," Diego Naranjo from European Digital Rights, a network of NGOs, told AFP.
Meta claims that user data is end-to-end encrypted before being fed into the RSC.
“Before data is imported to RSC, it must go through a privacy review process to confirm it has been correctly anonymised. The data is then encrypted before it can be used to train AI models,” it said.
(With inputs from AFP)
(At The Quint, we question everything. Play an active role in shaping our journalism by becoming a member today.)