It’s coming. Soon.

Meta will bring out its XR calling service with Pixel Codec Avatars (PiCA): a deep generative model of 3D human faces.

Unlike Apple’s Personas (which use front display and depth sensors), the Codec Avatars appear in 3D as full 360-degree photorealistic faces.

Let’s call it a “digital twin” for your face. As encoded virtual 3D models (using data captured by a smartphone), Meta will process these via AI models in the Meta cloud — powered by the new generation of AI computing.

This will allow integration into a virtual world (rather than simply appearing on a virtual screen or on a third-party device). On Meta’s website, you can see an example of the technology.

It’s the Holy Grail of video calling — and so many companies have come “unglued” chasing it. And because it’s the Holy Grail, people find it hard to believe it’s coming in the near future. There’s also the fact Meta has been working on Codec Avatars for years (at least since 2015). In the tech industries, there’s so much “news fatigue” with overhyped innovation that market research company Gartner makes a good living off of its graphic presentation of “the hype cycle.” The bigger the company, the more product ghosts that walk their corporate halls. And the more vaporware that flows through their hallways and meeting rooms.

We’ve heard about Codec Avatars for a long time … yet many of tech’s successes such as desktop publishing, IP-based video calls and machine language translation followed a similar pattern — years of public derision fed by unfulfilled promises and delays in bringing the actual products to market.

Meta’s Reality Labs’ previous work on Codec Avatars was clearly future-facing R&D. Now newly discovered job listings suggest Meta is getting closer to testing its immersive telepresence service with Codec Avatars among its own employees.

“You will be responsible for the highest level of polish, creativity, and interaction that allows the team to activate users and collect feedback that informs and empowers researchers and cross-functional partners toward the next generation of Codec Avatars,” says the job listing.

This Meta job listing also says the hire will be working “closely with engineers, scientists, and research product managers to build an internal XR calling service.”

The technology is now coming out of the lab and Meta will test it in its own internal communications.

One of the issues holding Codec Avatars back in the past was the complexity of scanning people to create their avatars. Lidar sensors now present in high-end iPhones and iPads might help to solve this, and the big investments in AI and AI PCs will certainly break through the scanning roadblock. Another Meta job ad explains how AI’s large-language models will pave the road for Codec Avatars.

Meta executives had previously said it may still take years for the company to bring Codec Avatars to its VR headsets. Yet, a VR news website discovered a recent Quest headset update already includes code to support Codec Avatar-powered video calls.

These “signs” lead some industry experts to think Meta will update us on Codec Avatars at this September’s Meta Connect conference. But if Meta is just launching an internal test at that time, it may be unlikely for us to have access until 2025.

The phone may not be ringing yet, but Meta is surely coming to call. And that call will not only “save face” for Meta’s Reality Labs but will also bring millions of faces to a new virtual life.