Apple of my iris, my shining Starline

Welcome!

As I prepped for Virtual Vector's launch, I thought about how long it might be until some new details about Apple's N301 headset would make headlines. Just a few weeks, it turns out.

Wayne Ma continued his streak of headset scoops at The Information on Friday with a report led by Apple's plans for iris scanning, a feature that could become the company's equivalent of Face or Touch ID for immersive headsets or glasses. Other headset differentiators discussed include downward-facing cameras to track the lower body, a weight less than that of Meta's Quest Pro, and support for magnetic prescription lens inserts.

Does any of this sound like a total slam dunk to you? The more we hear about its premium features, the easier it gets to believe that Apple really might launch this headset at $3,000. But, just to use iris scanning as an example, will the pitch of making it "easier for multiple people to use the same device" and securing payments be worth the expense and complexity at this time?

I might feel some of the bullish enthusiasm that many others have about Apple entering the market if we didn't know frustratingly little about its content plans. The hardware story is different. I don't doubt for a second that Apple is capable of releasing a headset that pushes today's boundaries–if anything, it's difficult to imagine Apple pulling the trigger if it hasn't locked in a few features it feels confident boasting about.

What I really struggle to get on board with is the idea that Apple's device itself will have anything uniquely meaningful in store. Over the past year we've seen Meta and others make big hardware moves in anticipation of competition from Apple, most recently with the reveal of the Quest Pro. Meanwhile, every detail that's been filled in around the story of Apple's pivotal standalone-or-bust moment has painted a picture of a headset design team that's facing the same challenges as everybody else. Apple could be the first to take a crack at solving certain headset woes, like the "alienation" issue, but there's little chance it's the only company driving toward a given approach.

We've now heard enough about the headset's features and development to have a decent sense of what's in store, and I'll happily eat my words if we're shown anything to the contrary. Now, though, it's clearer than ever that the truest "new" development when Apple launches might not be the hardware itself, but the ways the XR industry then contorts in reaction to it. Frankly, I can't imagine that story being overshadowed by one about specific sensors, weight, or resolution.

A triple from Google

Yes, Google recently made a point of signaling a new era by announcing plans to test AR prototypes in public, but notable XR news still comes out of Mountain View only sporadically. This week, though, Google has been busy: the company made three announcements that are worth a look.

• Project Starline, Google's line of immersive call booths kitted out with light field displays and numerous sensors, is moving into an early access program. So far, Google has identified Salesforce, WeWork, T-Mobile, and Hackensack Meridian Health as partners slated to get prototype booths soon. Google lists remote "employee onboarding and building rapport" as examples it has used Starline for internally, and I can see how those interactions might be improved by Starline's "magic window" effect. But it wouldn't surprise me if Google encourages or requires these test participants to stay away from tougher exchanges like performance reviews or–perish the thought–immersive layoffs.
• Tests of new Pixel phone-enabled features for business customers might convince you that Google Glass could have a real future. Sure, while Glass Enterprise devices still feature tiny peripheral displays like the failed Explorer Edition glasses did, the fact that they're intended just for hands-free information uses on the job already makes them more palatable. Add in live translation and transcription–an idea for glasses that piqued interest when Google recently pitched it with a sentimental spin–and maybe Glass starts to seem like less of a niche device and more like a legit communication tool.
• Google's Advanced Technology and Projects group (ATAP) gave a first look at what it's calling the ATAP Motion Platform, something that could tie into XR tracking and input solutions. In typical researcher speak, ATAP describes the platform as "a scalable tool to holistically treat motion data regardless of the sensor that it came from," a system that uses machine learning to capture and process human movements. That might make the platform useful for various approaches to body tracking in XR, but at one point in Google's video we also see a wrist-worn prototype letting a user control actions on a screen by making small hand movements. The parallel with Meta's research into neural input wristbands is striking.

Chasing the Horizon

On Saturday, Jeff Horwitz, Salvador Rodiguez, and Meghan Bobrowsky published a story for The Wall Street Journal that deserves a spot near the top of the pile of reporting on Horizon Worlds. Though the piece rehashes a few now-common tropes of metaverse coverage (it seems like Horizon's top comedy club may only be frequented by trolls and journalists), it more importantly presents some internal figures that help explain the shape Meta's immersive platform is in.

Remember around the time of the Super Bowl earlier this year when Meta confirmed that Horizon Worlds and Venues reached a combined total of 300,000 monthly users? The Journal reports that users have dwindled by over a third to less than 200,000, with Meta accordingly revising its year-end 2022 goal from 500,000 monthly active users down to just 280,000. That's bleak, especially considering that Meta can probably make a safe bet on thousands of first-time Quest 2 users giving Horizon Worlds a try this holiday season.

The story for pioneering creators in Horizon Worlds is also in a rough spot. According to documents viewed by the Journal, less than 1% of Horizon Worlds users end up making any worlds, with the top earning creation amongst them bringing in just $10,000 in payments. In total, tips for creators come out to a few hundred dollars.

None of this means that Horizon Worlds is doomed outright or that the platform can't be built upon moving forward. However, even apart from there being too many bugs and word that Meta employees themselves aren't using it much, stats like these are more than enough to make a "quality lockdown" for Horizon Worlds development seem like the right move. Call it whatever you like–measured, necessary, desperate–I'm sure it also wasn't an easy decision to make.

As for what else Meta can do to improve Horizon Worlds in the meantime, doing more to support the creators who have stuck with the platform so far couldn't hurt. I don't mean just with public shout outs or with help from Meta designers, but with more monetary support beyond what has been established so far. At the very least, by giving people more incentive to make and promote ambitious work in Horizon Worlds, Meta could further refine its sense of the platform's current strengths and weaknesses. The best case scenario would be that an injection of funds helps a noteworthy creator success story emerge in short order.

No matter how long something like that actually ends up taking, it seems Meta should be prepared to pump a lot more cash into jumpstarting Horizon Worlds than it initially planned to.

Connect closure

On Wednesday I posted my main takeaways from Connect 2022, but I promised to go a bit deeper on what Meta CTO Andrew Bosworth and Reality Labs Consulting CTO John Carmack had to say after the keynote presentation.

Bosworth hosted one of his regular Instagram "Ask Me Anything" sessions soon after Meta's keynote. One of the first questions he picked was "how will you win the fight against TikTok?" This, Bosworth responded, was "the wrong framing" for the ByteDance/Meta dynamic.

"TikTok has given us all a gift: they've taught us about a product that consumers want," Bosworth said, going on to talk about how he personally enjoys TikTok and welcomes the pressure to make Meta's short-form video offerings better. This was all familiar talk about innovation and competition, far from a new stance born of Meta's effort to turn the tides with Reels. But Bosworth dropped the "gift" talk in response to a question asking whether ByteDance should be allowed to compete freely in the US given that China restricts Meta.

"Boy, that's a very fair question, a very fair question," Bosworth said, leaving his answer at that. Could a move from ByteDance to push more Pico headsets in the US prompt Meta to agitate a little more forcefully for a favorable answer?

Later in the AMA, Bosworth shrugged off recent reporting on Horizon Worlds, characterizing the quality lockdown as "business as usual" for Meta's approach to development; a team pushes for features, it switches focus to bugs, rinse-repeat. This was before the Journal's report with the dismal user numbers, of course, which only makes it harder to take this nothing-to-see-here response at face value.

What really undercut these reassurances, though, were the negatives that Carmack focused on in his annual unscripted talk. While Carmack has had a reputation for not holding back in these sessions (remember his comments on the metaverse and "architecture astronauts" in 2021), this time around Carmack's whirlwind recap of the past year definitely swung toward a series of "grumpy" opinions and outlooks, as he put it.

Carmack did get his wish of hosting his talk in Horizon Worlds this time around, but he very quickly noted the lack of an audience and the fact that he was using a special version designed to keep the level of detail on his avatar constant (a prelude to everyone learning that Meta's demo of its new avatars was not done live). Later on, Carmack talked up his fondness for Meta's ongoing partnership with Qualcomm even as Meta works to develop its own silicon and aired objections to the effort to build a new immersive OS that was later dissolved.

"I can probably comment on that now," Carmack said. "We did have a large internal effort doing just that, and I was not supportive. In fact, one of my internal posts got reported and removed for not being sensitive enough."

It shouldn't be glossed over that Carmack went out of his way to address two of the bigger stories of turmoil inside Reality Labs from the past year, nor that Meta still seems to value his honesty and transparency these days. That said, if Carmack is ruffling feathers internally to the point where some of his critiques have been taken down, you've got to wonder if Meta will continue to let him share those thoughts so publicly moving forward.

More stories:

• Some early European orders of the Pico 4 headset will face shipping delays due to “unprecedented global demand," ByteDance said. [David Heaney / UploadVR]
• Viture announced that it has raised over $10 million in a Series A round led by BAI Capital and Verity Ventures. That comes in addition to $3 million raised through Kickstarter for its gaming-focused, $479 Viture One glasses. We’ll see if this round pushes Viture to take a more immersive XR path or if it’ll stick to making glasses that work as personal screens.
• Snap partnered with costume company Disguise to launch a Halloween costume AR shopping Lens. This seems like a good match for the current state of Snap’s try-on tech; you might not want to take a chance on buying an expensive garment you haven’t really worn, but skipping the trip to Spirit Halloween for something you’ll probably wear for a single weekend sounds OK. [Snap press release]
• Stress Level Zero's new game Bonelab made $1 million within its first hour on the Quest Store alone. [Justin Carter / Game Developer]
• Margaret Stewart, VP of Product Design and Responsible Innovation at Meta, announced that she is leaving in November. You can find Stewart's full post reflecting on her decade at the company here.
• Software engineer and Air Link torchbearer Amanda Watson also announced her exit from Meta after 7 years.
• Epson announced two new versions of its Moverio enterprise AR glasses. Note the difference in design: where the old Moverio models stuck to a glasses form factor, the new models can sport a halo-style strap or be clipped to a hardhat. More evidence that some AR end users really value being able to flip displays out of the way. [Epson press release]
• Hardware analyst Brad Lynch is planning to release a new "mega video" summarizing his findings and highly educated guesses concerning a standalone VR headset from Valve.
• Shiftall, the Panasonic-owned headset brand, unveiled an add-on for its prototype MeganeX VR headset that enables SteamVR tracking. The bulky module sort of ruins the look of the otherwise compact headset, and at present it's unclear exactly what support is planned for tethered and standalone VR with MeganeX. [Ben Lang / Road to VR]