IDCA NewsAll IDCA News
8 Nov 2022
Meta's Protein-Folding AI Serves as a Reminder That the Company Is More than a Metaverse Firm
Facebook, as the company was known before changing its name to Meta, has been working on something behind the scenes. They hinted at some new development during their recent conference call with investors, but they weren't specific. However, the company's latest press release reveals what they've been up to.
They've created an artificial intelligence that can predict how proteins will fold to help find cures for diseases like cancer and Alzheimer's.
Their research could lead to the discovery of hundreds or even thousands of potential drugs in less time than is currently possible. It's a major accomplishment for an augmented reality company, meaning we should expect more from them than we originally thought.
“It would be amazing if we could use machine learning to solve our world's most pressing problems,” said Jonas Hankin, a senior scientist at Meta who worked on the program. “But, after that, the sky's the limit.”
Earlier this year, Meta became a founding member of the AI Foundation. PyTorch was Facebook's creation, an open-source framework for deep learning.
Due to their increased focus on this metaverse technology, many people have speculated that their other ventures will suffer, causing their stocks to fall.
Meta shareholder Altimeter Capital's Brad Gerstner wrote a letter urging Meta to cut its metaverse investments and "solidify its position" as a world leader in artificial intelligence.
"Meta's investment in AI will lead to exciting and important new products that can be cross-sold to billions of customers. From Grand Teton to Universal Speech Translator to Make-A-Video, we are witnessing a Cambrian moment in AI. Meta is undoubtedly well positioned to help invent and monetize that future," wrote Gerstner.
The company's renaming to Meta may have caused the world to conclude that you were spending 100% of your time on Reality Labs instead of AI or the core business. Whatever the reason, that is certainly the perception."
The company released the ESM Metagenomic Atlas with over 600 million protein-folding predictions, as well as the underlying model and API of the software for general public use.
According to Meta, the language model of protein sequences could speed up structure prediction by up to 60 times.
"ESMFold shows how AI can give us new tools to understand the natural world, much like the microscope, which enabled us to see into the world at an infinitesimal scale and opened up to a full new understanding of life," explained Meta.
"Much of AI research has focused on helping computers understand the world in a way similar to how humans do. However, the language of proteins is beyond human comprehension and has eluded even the most powerful computational tools. AI has the potential to open up this language to our understanding."
Follow us on social media: