Artificial Intelligence Will Save the Metaverse
Facebook's plan to save itself is very much about obtaining AiSupremacy over the future iteration of the internet.
Hey Guys,
The hype of the Metaverse since 2020 has been incredible. What’s becoming increasingly clear is how central AI and AI Research at Meta is becoming to their attempt to create a consumer based VR world that could usher in new layers of the internet.
Do we need to artificial intelligence to build the Metaverse? Certainly the industrial Metaverse will augment companies with A.I. but what about the consumer version?
Perhaps we can find proof at Facebook’s attempt to build the Metaverse, that artificial intelligence is indeed a perquisite for this mythical AR/VR future of work play that some view as the next stage of the internet.
Meta will likely spend in the area of $100 Billion (or roughly $10 billion a year) to make it happen, where it might be said to be born around 2029.
Indeed if we view Artificial intelligence (AI) as the ability of a computer or a computer-controlled robot to perform tasks that are usually done by humans, how Meta is piloting voice-commands in the Metaverse, is truly an A.I. powered simulation.
AI is Having a Metaverse Moment
In Facebook’s eagerness to reincarnate itself and transcend the current internet, Supercomputers and build bots are helping them do it.
In February, 2022 Meta unveiled its AI Research SuperCluster (RSC) supercomputer, aimed at accelerating AI research and helping the company build the metaverse. The RSC will help the company build new and better AI models, working across hundreds of different languages, and to develop new augmented reality tools.
Meta is testing an artificial intelligence system that lets people build parts of virtual worlds by describing them that has a proof of the concept, called Builder Bot, could eventually draw more people into Meta’s Horizon “metaverse” virtual reality experiences. It could also advance creative AI tech that powers machine-generated art.
Do you enjoy my work? Then why not experience it all.
A.I. is also helping Facebook become Meta
Just as Google is trying to commercialize its expensive AI Lab, DeepMind, Meta is announcing a huge shift. Jerome Pesenti, Meta’s VP of AI for the past four years, will leave later this month as the company closer integrates his AI teams across various product groups rather than have AI function as a centralized organization.
In a blog on Meta called Building with AI Across all of Meta, Boz explained this wisdom. Andrew "Boz" Bosworth is currently CTO at Meta Platforms.
While it’s widely assumed ByteDance (from Beijing, maker of TikTok) is ahead of Facebook in A.I., Meta plans to double down on the technology.
The AI research team led by Yann LeCun, Meta’s chief AI scientist and an early pioneer in the field, will move to Bosworth’s Reality Labs, further consolidating resources in the AR / VR division that has already swelled to more than 17,000 employees.
AI is seen as a critical component of building the future hardware Meta hopes will define its vision of the metaverse. And, in the nearer term, it’s key to better competing with TikTok, which has popularized a new way AI is used to suggest content. (Outside of YouTube’s recommendation engine, TikTok’s is among the best).
Is the Metaverse Just an A.I. Generated Utopia?
In demos of builder bot and other Metaverse projects by Meta, Zuckerberg has described this as “all AI-generated.” In many ways we can think of the architecture of the Metaverse as an AI/VR symbiosis.
Whether for personalization or protection, whether to improve existing services or create entirely novel ones, Meta’s future including Facebook, WhatsApp and Instagram all depends on Meta’s ability to leverage the newest AI technology at scale.
In the recent Blog Boz is talking as if Meta were a AI-native and not just an Ad-tech company. Artificial Intelligence sits at the very heart of their work across Meta, he claims.
Teams like AI Platform, AI for Product, and more recently AI4AR have developed state of the art techniques – drawing inspiration from their colleagues at FAIR (Facebook AI Research) and counterparts across the industry – to leverage AI to improve our products, to better protect the people who use them, and to build innovative new applications.
Self-Supervised Learning at the Heart of Facebook’s Metaverse
AI is a critical component of building the future hardware Meta hopes will define its vision of the metaverse.
In many of the PR sessions over the last few months as Facebook become Meta, the idea that A.I. and self-supervised learning could transform VR with A.I. has becoming more clear. The leader of Facebook AI, Jérôme Pesenti, (who is stepping down) and co-managing director at Facebook AI Research, Joelle Pineau, have drilled into how Meta wants to unlock the metaverse with AI in a session titled “Unlocking the Metaverse with AI and Open Science.” Pesenti noted that AI is one of the keys to the metaverse.
Meta will leverage its new supercomputer and focus on A.I. among others things to unlock AI plans for a universal language translator, a new version of a conversational AI system, and an initiative to build new translation models for languages without large written data sets.
Self-Supervised learning is a bottleneck for building more intelligent generalist models that can do multiple tasks and acquire new skills without massive amounts of labeled data and Meta thinks it can be the holy grail to helping to manifest the Metaverse, the one that will save its company and the tarnished Facebook brand and legacy. Facebook even called it in a blog the “dark matter” of intelligence.
Meta said it believes that self-supervised learning (SSL) is one of the most promising ways to build such background knowledge and approximate a form of common sense in AI systems. But does this mean that AGI is needed to build the Metaverse? That remains dubious and unclear.
Self-supervised learning enables AI systems to learn from orders of magnitude more data, which is important to recognize and understand patterns of more subtle, less common representations of the world. This SSL is a predictive learning that will be able to anticipate our needs in the Metaverse better as our guide.
The general technique of self-supervised learning is to predict any unobserved or hidden part (or property) of the input from any observed or unhidden part of the input.
Meta’s New Organization of Artificial Intelligence
It seems to achieve the Metaverse, Meta has identified that its AI teams need to be more applied and work closely with the product teams.
In the new model Meta said it will distribute the ownership of these AI systems back to Meta’s product groups.
But Meta does this of course with the caveat that they must invest in a balanced portfolio that supports existing systems while also advancing the state of the art in AI.
Meta basically believe that this will accelerate the adoption of important new technology across Meta’s company while allowing them to continue to push the envelope. The teams tasked with driving AI advancements and best practices into the products they support will be known as AI Innovation Centers.
AIICs
Today (June 2nd, 2022) Meta’s announced the following changes:
The Responsible AI organization will join the Social Impact team.
The AI for Product teams that work to protect the people using our platforms, improve recommendations and make content more relevant, and improve our Ads and Commerce services will move to our product engineering team.
The AI4AR team will join with the XR team in Reality Labs.
Our AI research team, FAIR, will become a new pillar within Reality Labs Research. Its mission and charter will remain unchanged: Drive fundamental breakthroughs in AI through research excellence, open science, and broad collaboration.
In the future we might literally be at the mercy of the A.I. of companies like Microsoft, Google and Meta.
It’s doing to become even more powerful in our daily lives.
FAIR will continue to have incredibly strong leadership in place with Joelle Pineau, Antoine Bordes, and Yann LeCun. It was not readily apparent why Jerome Pesenti, Meta’s VP of AI is leaving.
Boz is also promoting (Montreal based) Joelle Pineau who will lead a new cross-functional AI leadership team.
Anyways guys I hope you found that insightful! It takes a lot of A.I. workers to make the Metaverse a distinct and distant possibility.
Join 72 other paying subscribers to support the channel and get access to a lot of premium content. This money goes directly into the work that I do and my basic needs and that of my family such as food, rent and household expenses. I cannot continue to write without community support.
Create your profile
Only paid subscribers can comment on this post
Check your email
For your security, we need to re-authenticate you.
Click the link we sent to , or click here to sign in.