Why Meta's Investment in Generative AI Is an Investment in the Metaverse
Ever since Meta published its Q4 2022 results in late January, speculation has run rampant on the future of the company's metaverse ambitions. The results showed the company's metaverse division, Reality Labs, had a $4.28 billion operating loss, bringing the total loss to $13.72 billion for the year.
A recent interview with Meta's CTO Andrew Bosworth in Nikkei Asia magazine sparked yet another round of the rumor mill, with some heralding it as the beginning of the end for Meta CEO Mark Zuckerberg's metaverse ambitions.
But rather than signaling a move away from the metaverse, the investment in generative AI is a sign of its next chapter.
Generative AI Can Fill the Metaverse Content Gap
Even the briefest analysis of what Bosworth said shows the company sees generative AI as a tool to help build and improve the metaverse. "So previously, if I wanted to create a 3D world, I needed to learn a lot of computer graphics and programming. In the future, you might be able to just describe the world you want to create and have the large language model generate that world for you. And so it makes things like content creation much more accessible to more people," said Bosworth.
And that's it in a nutshell. Rather than hampering the development of the metaverse, generative AI will facilitate its emergence in the enterprise.
Jeremy Bailenson, founding director of Stanford University’s Virtual Human Interaction Lab, put the relationship between the two into perspective. He said Meta has the one thing that many companies working in the virtual reality (VR) space do not have enough of: content.
He noted that VR hardware, such as Meta Quest 2 (formerly Oculus Quest 2), which is Meta's most advanced all-in-one virtual reality (VR) system yet, or Pico’s Neo 3, one of the first standalone headsets to challenge the Quest 2’s dominance, is already good enough to use at scale. What companies now need to do is focus on content.
“For the past few years, I have been successfully teaching in VR to over 500 students, each of whom spent hours per week in a headset for 10 weeks at a time," he told us. “Those two headsets each cost less than a smartphone and can render incredibly detailed scenes with low latency, high frame rate, high resolution, and a pretty good field of view."
Research from his Lab has confirmed the challenge with VR is content, not hardware. "There simply isn't enough to do in VR, as creating 3D content that works in a headset is incredibly difficult." The result, he said, is that generative AI may be one of the biggest contributors to adaptation of VR.
Currently if someone wants to build their own specific VR scene — for example a huge indoor room with picture windows and lounge chairs overlooking Lake Michigan — they would need to learn to use 3D Modeling software and/or program in Unity or Unreal, which is a huge barrier for entry and could take days even for those with the right skill set, Bailenson said.
“But generative AI could make that process as simple as typing the sentence above,” he continued. “3D scenes will be easier to generate with AI compared to 2D images of the same scenes."
He added that 3D content is largely symbolic, that is "subchildren" within scenes such as chairs have knowable separation from the other aspects of the scene such as floors, and information about size, position, and color of most objects within a scene are all typically part of the file format.
“Hence, by focusing on generative AI for 3D content, one is focusing on VR — think of all the content that Meta has access to in this respect,” he said.
Related Article: Facebook's Rebranding Embraces the Metaverse, But Not Everyone Is Convinced
Generative AI 'Extends' the Metaverse
Brandon Jung, VP of ecosystem at Tel Aviv, Israel-based Tabnine, argues that rather than being an impediment to Meta's metaverse ambitions, generative AI is a logical extension of it.
“Generative AI is the logical extension of the monetization of end user data that for Facebook today is through ads,” he said. "Facebook has a tremendous amount of proprietary data that they can use to build unique models, much like Google which has search, YouTube, and probably the most coveted codebase in the world.”
He points out that this maps very closely to Microsoft’s $7 billion purchase of GitHub, which then changed its terms of service to allow full use of the code with affiliates and gave it to OpenAI to build GPT-3 and Codex.
To compete in the space of generative AI, it is important to look for the pattern of large proprietary data sets, significant capital to train these foundational models, and a business model for distributing them, i.e. a public cloud, said Jung.
This is not the case for all generative AI use cases. For enterprise-specific use cases, like proprietary code, the solutions will need to be customized models that are small enough to run in the enterprise's datacenter or VPC and are tuned to their very specific patterns, he explained.
Related Article: Ready to Roll Out Generative AI at Work? Use These Tips to Reduce Risk
Where Apple's AR/VR Device Fits In
The wider market context is undoubtedly another factor in Meta’s generative AI push. While there are many companies in the VR or metaverse space, only a select few could pose a big enough threat to Meta to force it into action. Cupertino, Calif.-based Apple is one such company.
Learning Opportunities
The company is reportedly working on an AR/VR device which is expected to be revealed either at its Worldwide Developers Conference in June, or in September at its iPhone 15 event as a “One More Thing” moment.
For Apple boss Tim Cook, who has successfully launched the Apple Watch and AirPods, a successful VR/AR device launch and rollout could be the capstone on his legacy at the company and would carry him through until whenever he decides to retire, said Rolf Illenberger, founder and CEO of Munich-based VRdirect.
It would also increase the heat Zuckerberg already faces amidst a series of layoffs at Meta, despite having current market share in the device market via the Oculus.
Because it is Apple, Illenberger said, users should expect some sleek design and innovative features, as these devices will represent a major leap forward in augmented and virtual reality from what’s currently on the market. By bringing together the worlds of virtual and augmented reality, Apple is pushing the boundaries of what is possible and creating an entirely new way to interact with technology. From immersive experiences to virtual meetings and beyond, the possibilities are endless.
In fact, Illenberger believes an Apple announcement will mark a new milestone of marketplace maturity and reinvigorate some excitement as the “metaverse” hype wears off.
“There has been a fair deal of negative energy out of the tech world of late, but this has the potential to change that,” he said. “Apple has taken phones, headphones and watches and reimagined the user experience. I fully expect it to do the same for VR and VR devices.”
Related Article: AR and VR for Business: Where Immersive Technology Is Driving Real Value
The Metaverse Needs to First Succeed in B2B
It's still early days for the metaverse — and VR — and so it's not yet at a stage where adoption is the relevant measure for success, continued Illenberger.
“Many enterprises will be starting to roll out VR on a large scale, B2C adoption will follow in the next 1-3 years," he said. "This might not be what Zuckerberg (or Apple) is stating publicly, but everyone in the industry knows, that you will first have to win the B2B game, before making this a B2C business."
He added that the current hype around generative AI and the metaverse has a positive impact in that it puts the emphasis on the quality of experience instead of the quantity of experiences offered, he continued. It also takes pressure off Apple to get something out the door, allowing it to cater to its high-end consumers and community of developers willing to work around a new device and its flaws.
The result, he said, is it puts the emphasis on successful VR and AR B2B use cases as opposed to mass-selling a product consumers may or may not be ready for.
Full Steam Ahead on the Metaverse
None of the above accounts for the ongoing metaverse developments at Microsoft and Google, nor the dozens of smaller companies working in VR, the metaverse and generative AI.
All of this adds context to Meta's current moves and we can expect to hear similar ones from its competitors in the not-too-far-off future.
At the end of the day, given the company's strong roots in AI, a move into generative AI is a logical step, regardless of the metaverse. But the clear symbiosis between the two suggests the company is doubling down on its metaverse commitments.
About the Author
David is a European-based journalist of 35 years who has spent the last 15 following the development of workplace technologies, from the early days of document management, enterprise content management and content services. Now, with the development of new remote and hybrid work models, he covers the evolution of technologies that enable collaboration, communications and work and has recently spent a great deal of time exploring the far reaches of AI, generative AI and General AI.
Featured Research
Buyer's Guide
Read Now
Research Report
Read Now
eBook
Read Now
Research Report
Read Now
White Paper
Read Now
Research Report
Read Now
Guide
Read Now
Guide
Read Now
Related Stories
Reworked CONNECT 2024 - Pre-Registration is Now Open!
