How Revolutionary Are Meta’s AI Efforts?
Mark Zuckerberg introduced Facebook’s rebranding to Meta at the company’s annual Connect event last year to reposition the company for the “new internet,” the metaverse.
The metaverse has been around for some time as a kind of urban legend, perhaps aptly described in the 2011 science fantasy book Ready Player One. That was until some of the biggest names in tech started investing heavily in related technologies, including virtual and augmented reality (VR/AR), Internet of Things (IoT), and artificial intelligence (AI).
Today AI is one of the most exciting technology fields to work on. Zuckerberg said the metaverse is something he’s wanted to work on since even before the conception of Facebook. And the company’s Meta AI research lab is on the cutting edge of both AI and the metaverse.
The new direction was met with some trepidation. The social media site Facebook, which retains its name, is already infamous for its black-box algorithms, which are going to grow more and more complex under the Meta AI efforts.
While this news is undoubtedly exciting for enthusiasts and futurists, what are the practical real-world implications beyond the metaverse hype?
See the Top Artificial Intelligence (AI) Software
How Far Along is Meta AI?
Despite the initial excitement, Meta AI is years away from reaching its ultimate goal of having a fully operational metaverse. Meta AI is Meta’s longest scheduled project, perhaps as big in development time as it is in ambition. But that doesn’t mean it can be ignored, as it has many partially developed applications.
Meta is presently leading the AI race, with many of its applications in direct competition with giants across different categories. Meta is working on a voice interface like Apple’s Siri or Google Voice Assistant. Meta is also competing with VR gaming consoles such as Hololens with the acquisition of Oculus.
Additionally, some of Meta’s projects include an AI-based image tagging and image search algorithm that was shown to beat the FBI’s image recognition ability at the IEEE Computer Vision Conference. Meta AI can also generate text predictions in messages like Google assistant through DeepText.
Meta AI Controversies
Facebook has been guilty of a number of ethical failings over the years and has been called out for prioritizing profit over safety by Frances Haugen, a former product manager at the company. On more than one occasion, it has been reported that this prioritization has led to the spread of misinformation and hate speech, which tends to engage more people, and the company has been accused of ignoring the negative effects of social media on teens and other groups.
A huge chunk of Meta revenue comes from Facebook ads. However, the ads are only valuable if the content you find on the site is engaging. Reports suggest that Facebook algorithms unilaterally favor engagement and, as a result, might promote misinformation and hate speech.
In response to these claims, Meta announced several changes in June 2022, including:
- The Responsible AI organization is set to join the Social Impact team.
- The AI for Product teams working to protect users on any of the Meta-owned platforms will move to the product engineering team and focus on improving recommendations and making content more relevant, including ads and commerce services.
- The Meta AI research team, FAIR, will join the Reality Labs Research to serve as foundation and support, with a mission to drive breakthroughs in AI through “research excellence, open science, and broad collaboration.”
The Latest Developments in Meta AI
Meta has made the news multiple times since its rechristening, and not all of it has been troublesome. Just this May, Meta announced that it has created a massive new language model that will be freely available to researchers around the globe.
In a move that is not typical of a for-profit organization, Meta surprised its followers, saying that the idea was “democratizing access to large-scale language models, to maintain the integrity and prevent misuse.”
This will be the first time a fully trained language model of this size and scope will be accessible to researchers. The move has been well-received by critics of privately owned and funded research on AI models.
Meta has been full of surprises this year, releasing its AI Research SuperCluster (RSC), the world’s fastest AI supercomputers, earlier this year. The supercomputer will accelerate AI research and help Meta build the metaverse. New announcements from the AI effort come frequently, the latest an AI chatbot.
Also read: Python for Machine Learning: A Tutorial
What Areas Will Meta AI Influence?
What influence might Meta AI have on the evolving virtual world of the metaverse?
Marketing
Your meta life could involve digital clothing and world-building, and marketers would have to consider it while strategizing on sales through the metaverse. Game developers and marketers are no strangers to in-game marketing and have displayed custom skins, virtual locations, in-game advertisements, and promotional items. The metaverse would be no different.
Culture
The metaverse brings together an alternate online subculture pushed forward by gamers, buyers, and the brands serving them.
In December 2019, GTA V allowed players to dress up as protesters in Hong Kong. Gamers put on black clothes, with yellow hard hats and gas masks to start a riot in-game. Chinese players reciprocated by dressing up as the police and fighting back.
The GTA incident was a unique phenomenon, showing unexpected ways people might use the metaverse. People from different cultures can unite as groups driven by a shared sentiment, creating a unique cultural phenomenon.
Economy
The metaverse is the advent of a shared, virtual economy. People might behave differently in the metaverse than in real life, creating alternate spending patterns.
Whenever the metaverse rolls towards global adoption, people from different economies will be on a shared platform. The metaverse might facilitate in-game trade. Brands might turn into pseudo-super-powers. Only time will tell how a metaverse might affect what we know about economics.
Why We Need to Take Meta AI Seriously
If it is not Meta AI, it could be any other AI project, but the virtual world platform race is on. What makes Meta unique is its application-oriented approach. Meta has the benefit of being the parent company to the world’s most prominent content churning machine, which allows Meta to leverage user-generated data to build its AI.
However, there don’t seem to be any signs of significant disruption, and growth appears to be moving slowly and steadily. So when someone says Meta AI is a big deal, it is so in scope but perhaps more evolutionary than revolutionary. Only time will tell how world-changing it will be as we wait for more updates from the social media – and now AI – giant.
Read next: 10 Top Data Companies
The post How Revolutionary Are Meta’s AI Efforts? appeared first on IT Business Edge.