Alex Heath, reporting for The Verge:

The black Clark Kent-esque frames sitting on the table in front of me look unassuming, but they represent CEO Mark Zuckerberg’s multibillion-dollar bet on the computers that come after smartphones. 

They’re called Orion, and they’re Meta’s first pair of augmented reality glasses. The company was supposed to sell them but decided not to because they are too complicated and expensive to manufacture right now. It’s showing them to me anyway.

I can feel the nervousness of the employees in the room as I put the glasses over my eyes and their lenses light up in a swirl of blue. For years, Zuckerberg has been hyping up glasses that layer digital information over the real world, calling them the “holy grail” device that will one day replace smartphones…

Orion is, at the most basic level, a fancy computer you wear on your face. The challenge with every face-computer has long been their displays, which have generally been heavy, hot, low-resolution, or offered a small field of view.

Orion’s display is a step forward in this regard. It has been custom-designed by Meta and features Micro LED projectors inside the frame that beam graphics in front of your eyes via waveguides in the lenses. These lenses are made of silicon carbide, not plastic or glass. Meta picked silicon carbide for its durability, light weight, and ultrahigh index of refraction, which allows light beamed in from the projectors to fill more of your vision.

Orion is an incredible technical demonstration, but it’s only that: a demonstration. It’ll never ship to the public, by the admission of Mark Zuckerberg, Meta’s chief executive:

Orion was supposed to be a product you could buy. When the glasses graduated from a skunkworks project in Meta’s research division back in 2018, the goal was to start shipping them in the low tens of thousands by now. But in 2022, amid a phase of broader belt-tightening across the company, Zuckerberg made the call to shelve its release.

There’s a reason Orion will never truly come to the market anytime soon: it’s technically impossible. Just to make this ultra-limited press product, Meta had to put the computer in a separate “wireless compute puck,” which connects via Bluetooth to the main glasses. It also couldn’t master hand tracking, which is supposed to be the primary method of input confirmation, so it made an electromyography-powered wristband to “interpret neural signals associated with hand gestures,” in Heath’s words. All of this costs money — and no small amount. Even if Orion were priced at $10,000, it would just be too expensive and technically impossible to ever be mass-produced in any quantity. Every Orion device is evidently handmade in Menlo Park with love and kisses from Zuckerberg himself, or something similar.

But if all one did was watch Meta’s hour-plus-long Meta Connect annual keynote from Wednesday, that wouldn’t be apparent. Sure, Zuckerberg made clear that Orion was never meant to ship, yet he didn’t position it like the fragile prototype it truly is. The Orion glasses Heath — and seemingly only Health and a few other select members of the media — got to try are as delicate as a newborn baby. They’re not really a technology product as much as they are the beginning of an idea. Without a doubt, I can confidently say Apple has an Orion-like augmented reality smart glasses prototype running visionOS in Apple Park, but we won’t get a look at it until five or six years from now. I keep hearing people say that Meta just killed Apple Vision Pro or something, but that’s far from the truth — what we saw on Wednesday was nothing more than a thinly veiled nefarious attempt to pump Meta’s stock price.

Zuckerberg, in a pregame interview with The Verge, said he believes an Orion-like product will eventually eclipse the smartphone. That’s such an outlandish claim from someone who didn’t even see the smartphone coming until 2008. What’s better than a finicky AR glasses prototype with low-resolution projectors and thick frames? A compact, high-resolution, gorgeous screen, lightning-quick processor, modem, hours-long battery, and professional-grade cameras all packed into one handheld device. A mirrorless camera, a telephone, and an internet communicator — the iPhone, or the smartphone more broadly. People love their smartphones: they’re discreet, private, fast, and easy to use. They don’t require learning gestures, strap-on wristbands, or connecting to a wireless computer. They don’t require battery packs or weighty virtual reality headsets with Persona eyes. From the moment it launched, the iPhone was intuitive and it continues to be the most masterfully designed piece of consumer technology ever made.

No glasses, no matter how impressive a technical demonstration, will ever eclipse the smartphone. No piece of technology will ever be more revolutionary and important. These devices can and will only reach Apple Watch territory, and even that amount of success isn’t inevitable or to be taken for granted. They’re all auxiliary devices to many people’s main computer — their phone — and that’s for good reason. I’m not saying there’s no purpose for so-called “spatial computing” in Apple parlance, because that would be regressive, but that purpose is limited. There’s always room for new computing devices so long as they aren’t stupid artificial intelligence grifts like the Humane Ai Pin or Rabbit R1, and I think some technology company (probably Apple) will eventually succeed in the spatial computing space. As Federico Viticci, the editor in chief of MacStories, says on Mastodon, soon we’ll all be carrying around an iPhone, Apple Watch, and Apple Glasses. I genuinely see that future in just a few years.

But in the meantime, while we’re waiting for Apple to sort out its Apple Vision Pro conundrum, we’re stuck in this weird spot where Mark Zuckerberg, of all people, seriously thinks he’s game to talk down Apple and OpenAI. The truth is, he knows nobody but some niche developers care about his Meta AI pet project; all eyes are on OpenAI. No matter how much he tries to shove his chatbot down people’s throats on Instagram, they’re not interested. He’s gotten so desperate for AI attention that he’s resorted to inserting AI-generated images in people’s Instagram timelines, even if they don’t want them. One day, Instagram’s going to turn into an AI slop hellscape, and this is the supposed future we’re all expected to be excited about. Zuckerberg’s strategy, in his words, is to “move fast and break things,” but in actuality, it’s more like, “Be a jerk and break everyone else’s things.” Zuckerberg is fundamentally an untrustworthy person, and his silly Orion project deserves no more attention than it has already gotten. Just don’t forget to pay your respects to Snap’s grave on the way out.

Now, back to reading the tea leaves on this OpenAI drama. Sigh, what a day.