Apple Vision Pro In Depth: FAQ and Industry Implications
Avi Greengart, Lead Analyst
In the weeks since Techsponential got heads-on with the Apple Vision Pro spatial computer (Apple’s term) at WWDC23, we have been asked numerous questions from journalists, Apple partners and resellers, and Apple competitors. This report provides a more comprehensive look at the hardware, software, ecosystem, and implications of Apple’s entry into AR and VR.
Framing: Is Apple Vision Pro a VR headset?
During the WWDC keynote, Apple called the Vision Pro a “spatial computer” that has “environments” and “levels of immersiveness.” Apple never used the terms, “VR,” “AR,” “MR,” “XR,” or “metaverse” publicly or in my conversations with Apple executives. Apple often tries to brand industry terms as Apple-specific, and there’s definitely some of that going on here, but this is more about Apple redefining VR as a computing platform rather than a technology. “VR” and “AR” are clunky terms, especially if you mix the technologies. To get around this, Microsoft used “mixed reality” and Qualcomm “extended reality” but those terms leave out fully immersive virtual reality experiences. In the end, “spatial computing” is both more expansive and more descriptive of what’s really going on. It also allows Apple to claim to be at the forefront of a new computing platform, which is a bit of a marketing coup. Bottom line: I expect the term “spatial computing” to catch on.
Core Principles: What Makes the Vision Pro Unique?
With a few exceptions, Apple is not first to market, but when it does enter a market, it typically does so with a different approach and user interface. The Lisa and Mac were the first commercial PCs to use a GUI. The iPod was tethered to iTunes and controlled by a clickwheel. The iPhone was a multi-touch mobile computing device, and its potential was realized a year later with the App Store. The iPad was a large multi-touch iPod rather than a pen-based laptop. We have had modern VR headsets around for over a decade; what is Apple doing differently?
Retransmit for AR – While it is not hard to find use cases for VR (simulation, design, gaming, entertainment), the potential for augmented reality is nearly endless. Apple’s CEO Tim Cook has publicly touted the value of AR for years. However, Apple appears to believe that the technology for see-through AR glasses is still years away; the field of view is too small, and the ability to create opaque digital objects is too limited. Apple’s solution is build a VR headset that recreates your real-world environment and then allows the user to adjust how immersive the virtual world is. This can range anywhere from adding from digital objects (complete with shadows) to your field of view, all the way to completely digital environments.
AR Drives Hardware (and Pricing) Requirements – Once Apple decided to go all-in on camera pass-through for AR, that necessitated eye-level resolution so that your view of the real world is clear. It also required incredibly low latency so that movement of your head doesn’t conflict with what your eyes see and cause nausea. To manage all those inputs and pixels requires custom silicon (Apple’s new R1), perfectly calibrated external cameras, and lots of CPU, GPU, and NPU power. Although Apple couldn’t get the technology to fit in a pair of glasses, AR experiences require as much mobility as possible, dictating that the Apple Vision Pro be (almost) entirely self-contained. The headset is not tethered to a computer, it is a computer, with the same Apple M2 processor found on Macs. (The one design problem Apple couldn’t solve is electrical power; the Vision Pro still requires a battery or a long cord to a wall outlet.) Those high-end displays, custom chips, cameras, and sensors are expensive, and Apple has never been willing to lose money on hardware to build a market, necessitating a $3,500 price point before additional customization and storage levels.
New Computing Paradigms Call for New Interface Methods – VR and AR headsets have an input problem. Most AR glasses have resorted to buttons and swiping motions at the temple that are limited and clunky. Most VR headsets use handheld controllers; these provide better control – and can be great for some types of games – but it means that your hands are full. It also means that you have two more things to keep charged and find every time you start a session. Microsoft tried gaze and gestures on the HoloLens, but the technology that HoloLens uses is imprecise and adds frustrating delays throughout the interface. Apple’s eye tracking and gestures system on the Vision Pro is exceptionally refined: it takes seconds to learn. It leaves your hands free. It provides developers with a flexible, precise, and natural-feeling way to manipulate things in 3D space.
Reducing VR’s Inherent Isolation – Putting on a headset that blocks your vision isolates you from your environment, including any objects or people in it. Apple does three things to address this issue: Apple allows the wearer to adjust how much of the digital world intrudes on the retransmitted view of the real world; when you are fully immersed, nearby people and objects fade into view, allowing you to interact with the people and avoid tripping trip over the coffee table; and Apple provides social cues to the people around you as to what you’re doing by digitally rendering of your eyes on an lenticular OLED display on the outside of the Vision Pro. This last aspect was not demoed live; it is hard to predict how well it will work or be received.
Ecosystem Advantages -- Apple’s comprehensive software, services, and devices ecosystem means that the Apple Vision Pro will be useful out of the box: most iOS and iPadOS apps will run as virtual resizable-anywhere windows. AR apps written with Apple’s AR Kit for iPad have a headstart on spatial computing aspects of the headset. Apps written for Unity can be brought into Apple’s development environment. Mac users can treat the Apple Vision Pro as a private virtual workspace. There are also less obvious aspects to Apple’s ecosystem: Apple can use iris scanning and eye tracking as core parts of the Vision Pro experience without significant privacy fears because it is not primarily selling advertising; when Apple says that your gaze information is not discoverable or sold to third parties, it is credible. Apple also plans to use its retail stores as a way to show off the Apple Vision Pro and set up appointments for sizing and customization.
Criticism: Is the Apple Vision Pro Too Expensive?
Leading up to its launch, rumors pegged Apple’s headset at $3,000. When the actual price of $3499 was announced, there were gasps in the audience at WWDC, and I even heard some cries of, “nooooo!” This is not going to be a casual purchase. In fact, most buyers will almost certainly pay significantly more for the Apple Vision Pro for more storage, better straps, a case, batteries, and custom correctional lenses. The Apple Vision Pro is more capable than other VR and AR headsets, but Apple’s pitch as a computing platform, entertainment system, and multi-monitor setup will not resonate with regular consumers at launch. This is a 1.0 product aimed at early adopters and businesses who can afford to spend high-spec MacBook Pro money on something new. In the short term, this is not a problem. Apple does not need to sell tens of millions of Vision Pros at launch – in fact, it can’t manufacture or sell that many, given the noise from the supply chain about yields and complexity, along with the bespoke sales process the Vision Pro requires. Apple will need to provide regular consumers with a reason to spend $3500 on spatial computing at some point (and/or find a way to get the price down), but demand for the Apple Vision Pro among enthusiasts is expected to exceed supply for a while.
For enterprises looking for a platform to build AR apps, the Apple Vision Pro is actually priced competitively at launch, given its capabilities and the developer resources available. However, launch price is not too much of a consideration given the timescale of enterprise development. Businesses will mostly be buying individual units for their IT and R&D staff in 2024 and 2025 with larger scale deployments not starting until 2026 and beyond.
Criticism: Is Zuckerberg right?
One of the sharpest critiques of Apple’s Vision Pro came from a direct competitor. After noting that Meta had already considered all of the technology incorporated into the Vision Pro, CEO Mark Zuckerberg complained in a video to Meta employees that all the Vision Pro demos were of individuals disconnected from others and not being active.
“Our vision for the metaverse and presence is fundamentally social. It’s about people interacting in new ways and feeling closer in new ways. Our device is also about being active and doing things. By contrast, every demo that they showed was a person sitting on a couch by themself. I mean, that could be the vision of the future of computing, but like, it’s not the one that I want.” - Mark Zuckerberg
Zuckerberg is at least partly correct. Meta does not have Apple’s silicon division, but Meta certainly evaluated the core technologies for eye tracking, hand tracking, and pass-through that Apple used in Vision Pro. However, whatever Meta’s engineers may have considered, Meta did not build a Vision Pro, so the fact that Apple did not break the laws of physics does not help Meta’s competitive position one bit.
The second part of the critique is more interesting. On the social front, it is true that Apple did not show off the gaming and exercise apps that are the best sellers on Meta Quest, though there is no reason that developers couldn’t build them. There may be some reluctance to use a $3,500 headset for musical lightsaber routines, but on a technical level there is no reason why Beat Saber won’t be among the first ports to Apple Vision Pro using gestures rather than controllers. There do appear to be limitations to how much you can move around – the developer tools Apple has released point to a 10x10 active area, so commercial multi-room VR gaming does not appear possible with version 1.0. However, my own demo involved getting up off the couch, walking around the room, and interacting with virtual butterflies and dinosaurs on the far side of the room. There are multi-user games for the Meta Quest 2 like Eleven Table Tennis, but there is no reason to think this type of game is not possible with the Vision Pro. Whatever its limitations, Apple Vision Pro is not solely a stationary experience.
The notion that Quest is social and Vision Pro is isolationist is fundamentally backwards: Apple uses much of the technology in the Vision Pro to reduce the sense of disconnection from your environment. The user can see out of it, and others in the room can see a facsimile of the user’s eyes to know if they are immersed. Even if the user is fully involved in a virtual environment, moving close to them will fade you into their view. In contrast, when you put on a Meta Quest 2, you are blocking yourself off from the real world. The Quest 3 will likely have better pass-through capabilities when it launches later this year, but not at the level of the Vision Pro.
Meta has apps for collaborative work, but they are not widely used, and Apple has already demonstrated FaceTime inside the Vision Pro. The experience as demonstrated to me was good enough to make Meta’s cartoon avatars look silly, but still not nearly as effective as seeing a video of someone’s actual face. Meetings in VR are great when collaborating on a digital object (or a digitized physical object), but for anything else, today’s VR meeting tech just gets in the way. Combine this with the fatigue factor of wearing a headset, and people who spend all day in meetings are likely going to prefer face to face or Teams/Zoom.
Apps and Developers: Can Apple catch up with Meta on applications?
Catch up? Developer engagement is one of Apple’s superpowers. Apple is already ahead. Meta recently said that it has over 500 games and apps for the Quest platform, while Apple has tens of thousands of iOS and iPadOS apps that will run in the Vision Pro as 2D windows at launch. Apple and Disney will have streaming content, and it will be shocking if other streaming apps don’t follow suit. Microsoft has committed to porting its Microsoft 365 (Office) productivity suite to Vision Pro, and I expect that it will be bringing its Dynamics enterprise software to the platform as well. It is not clear how many native VisionOS apps will be available at launch, but the tools and APIs are similar to ARKit, and Apple is enabling Unity developers to adapt content for Vision Pro. The question isn’t whether Apple Vision Pro will have apps, but what types of apps beyond phone/tablet apps it will have, and how compelling they will be.
Apps and Developers: How will the Apple Vision Pro handle games without controllers?
Gaming is big in VR. Games are a key reason to buy a Meta Quest 2 and the only reason to invest in a PSVR 2 on top of a Sony PlayStation 5. It seems likely that entertainment will be a key reason for consumers to buy an Apple Vision Pro as well. However, Apple is not shipping a physical controller with the Vision Pro, and, given how effective its eye tracking and gesture navigation system is, developers should not expect Apple to change that. It’s a bit like expecting Apple to put a physical keyboard on the iPhone just because that’s how input was done on smartphones up to that point.
At the WWDC keynote and accompanying demos, Apple showed off 3D movie watching, streaming content, and an interactive dinosaur experience. It did not show off any other game types, not even the type of games that have been successful in AR or VR – nothing from Niantic, no FPS or open world exploration games, no racing or flying simulators, and nothing from the rhythm exercise or fitness categories. I do expect Apple to create Apple Fitness programs for Vision Pro – it certainly should not expect Meta to port its own fitness app, Supernatural, over to a rival.
Some games will need dedicated controllers to be viable in VR; you need a wheel and pedals for driving games and a yoke for flight simulators. However, beyond that, I expect that we will see a lot of experimentation with immersive, interactive entertainment experiences that use your hands and gestures rather than a game controller. Mobile games do not use controllers, and mobile gaming is nearly twice as large as console and PC gaming combined. This does not mean that gaming on Vision Pro is doomed, just that it is different – and if the games are good enough, that difference may be the reason people buy a Vision Pro. Personally, I’m hoping that someone steals the best idea in Ready Player One (the book, not the movie): go inside scenes from movies and get graded you on how well you act out the part. Forget Beat Saber, why can’t I pretend to be an actual Jedi or Sith inside the movie? Disney, please take my money.
Using Apple Vision Pro: Is Taking Spacial Photos Dystopian?
One of the use cases that Apple showed off in the keynote was a father capturing images of a birthday party while wearing the headset and then reliving that experience. I got to try playing back 3D spatial photos and videos during my demo, and it’s a seriously mind-bending experience – more akin to something out of Harry Potter than browsing through Photos.
However, highlighting this use case was a mistake on Apple’s part. First, to understand the value of spatial photos and videos, you need to experience them. More importantly, there is no way that consumers will – or should – be comfortable taking those images and videos in the first place, at least not via the headset. It’s already bad enough that people interject a phone between themselves and the landscape, family members, and concerts that they attend; putting on a headset, even with pass-through and rendered eyes on the outside completely removes you from the actual experience in favor of capturing a semblance of it for later.
Apple could add the ability to create spatial media in future iPhones and iPads, and that would neatly solve the problem. Apple can’t talk about that now, so it should have left this use case out until it has handheld capture devices to pair with the Apple Vision Pro’s playback. In the meantime, panoramas taken on existing iPhones are absolutely stunning when displayed room-size around you inside the headset.
Using Apple Vision Pro: Is Battery life sufficient?
The short answer is no, two hours of battery life without the ability to hotswap the battery is not sufficient for all- or even partial-day use. Apple appears to be allowing the Vision Pro to be tethered to a wall outlet (or possibly to a Mac), and it may be possible to daisy chain multiple batteries together.
That said, taking a break after two hours of use is likely going to be recommended regardless. My demo lasted between 30 and 40 minutes, and while I did not experience any eyestrain or motion sickness, my sense was that longer sessions would cause at least some eye discomfort, at least for analysts who are old enough to have covered the original iPhone launch.
Using Apple Vision Pro: Connectivity
The Apple Vision Pro ships with WiFi. A cellular option is almost certainly on Apple’s wish list, but not likely on its roadmap until battery life can be improved and Apple expands the Vision Pro’s usage area from limited physical spaces to, well, everywhere else. Apple also wants to avoid paying a Qualcomm tax, so a cellular Apple Vision Pro will also hinge on Apple’s internal modem development efforts. In the meantime, carriers should be instructing their enterprise sales and consulting divisions to bundle wireless 5G hotspots with Apple Vision Pro for use in the field.
Competitive Landscape
The VR market has been just around the corner for decades. Meta renamed itself and invested tens of billions of dollars a year trying to create a new platform that it can monetize without interference from Apple or Google. However, Meta hasn’t really figured out what people want to do in VR. Its shared work environments aren’t better than using Zoom, adoption has been low, and adding legs to avatars will not change that. VR has been moderately successful in gaming (driven mostly by small independent development shops) and exercise. Sony has sold millions of PSVR’s and hundreds of thousands of PSVR 2’s; Meta has sold over 20 million Quest 2’s. However, even in consumer VR, sales growth has stalled. Enterprise AR and VR is used for 3D design, simulation, and training, but many trials concluded without resulting in significant deployments. In the post-pandemic tech purge, investors and executives concluded that if VR didn’t go mainstream during the pandemic, it wasn’t likely to do so any time soon, and they pulled back on anything metaverse to focus investment on generative AI instead.
Apple’s entry into spatial computing changes that. When Apple announced the iPhone in January 2007, Google immediately switched from creating a BlackBerry clone with its first Android phone to a more touch-centric, app-oriented experience. The introduction of Apple Vision Pro is almost certainly causing similar reversals at hardware, software, and content companies today. One early example is Disney, which shut down its VR division just a few months ago, only to appear at Apple’s keynote as a day one partner.
Rivals will have plenty of time to respond, but they do need to start now. The first Apple Vision Pro won’t reach the market until 2024, it will cost a small fortune, and its capabilities and comfort will limit its utility. However, this is a playbook Apple has perfected: launch an initial product with a unique approach to a stagnating market, tie it into a robust ecosystem, then iterate, iterate, iterate. By version two (2025) or three (2026), the supply chain will have had time to ramp up, the use cases will have crystalized, and Apple will have addressed critical comfort, battery, and price points. Beyond that point, material science and components will start enabling Apple to move towards more AR-first experiences. That puts 2027 as the tipping point for catching up to Apple or potentially getting stuck trying to compete in a market where Apple dominates hardware sales, maintains full ecosystem control, and monopolizes industry profits.
2027 is still several years away, but those intending to compete with Apple need to move to intercept Apple today. There is nothing on the market today (or that I have seen in development for the near future) that approaches the capabilities of Apple Vision Pro version 1.0. There are three potential response approaches:
Build the VR market from the bottom up. First, find a sweet spot in VR where today’s technical capabilities are good enough to provide value at a mainstream price point. Next, grow the product’s capabilities over time to match what a lower-priced future version of Apple Vision looks like a few years out. This is essentially what Meta has been doing with its Quest line and ByteDance has been doing with Pico. Fitness and short gaming sessions have emerged as key app types for these headsets, but sales momentum for the $400 - $600 headsets has stalled, and developer interest is strongest among smaller shops. Perhaps Meta’s Quest 3 will reverse this trend when it launches this fall, but if it doesn’t, that will be a problem. This build-up-from-below strategy depends on growing a huge installed base and developer ecosystem before taking a portion of that audience upmarket.
Redirect mid- and high- priced VR systems to compete directly against Apple Vision Pro. Tim Cook likes to say that, “only Apple could have done this.” When it comes to the computing capabilities and price point of a VR headset, even competitors agree that this is accurate – current “pro” headsets from Meta, Valve, and HTC Vive have not been willing to push beyond the $1,000 - $2,000 range. If Apple has raised the ceiling on VR headset pricing, could rivals build a Vision Pro equivalent at or below the same price?
One challenge will be silicon. Apple is putting an Apple M2 in the Vision Pro, along with a custom R1 chip for I/O, and the M2 isn’t even the most powerful chip in Apple’s silicon arsenal. Vendors have been pushing Qualcomm to keep pricing down on its XR chip line, but now that there is a clear benchmark to hit, there should be a chorus of vendors asking for Snapdragon XR chips that can handle extremely high resolution pass-through, have Qualcomm’s upcoming Oryon CPU cores, and plenty of GPU space on the die.
Assuming that Qualcomm is able to deliver (not a bad assumption, though the timing won’t be immediate), it won’t change fundamental supply chain constraints around TSMC capacity and eye-resolution displays. Competing directly against Apple also requires building up a developer and app ecosystem that far exceeds what Valve and Meta have today. ByteDance might actually be in a better position as the Chinese market balkanizes by focusing most of its energy at home. Google is a wildcard.
It’s also worth noting that there have been exceptions to the VR price ceiling before Apple’s entry. Microsoft’s HoloLens 2 debuted at $3,500, and Varjo’s XR3 costs $6,500 before $1,500 annual software licenses, plus a $4,500 PC required to drive it. Sadly, outside of fulfilling its Army contracts, Microsoft has seemingly given up on AR and VR hardware. (Microsoft is all-in on software). XR is Varjo’s entire business, and it is actually fairly isolated from Apple in the short term thanks to a price-insensitive customer base who use Varjo’s headsets to do things that Apple’s Vision Pro is unlikely to be capable of at launch. Varjo should expect that some of its design market will shift to Vision Pro in the 2025 timeframe (or whenever Apple can build its headsets in volume), but auto and aerospace simulation may require significantly more GPU horsepower than the self-contained Vision Pro can muster, even with an M2. There are regulatory barriers to entry as well. If Varjo pushes deeper into this market (and improves its own comfort and pass-through), it can comfortably inhabit its niche markets even as Apple makes it difficult to push down into lower price points with products like the Aero.Start with AR. Apple’s implicit assumption with the Vision Pro is that you can’t do AR today, so it is starting with VR pass-through and working its way towards AR. What if Apple is wrong? It isn’t possible to build Tony Stark’s glasses today: you’d need an order of magnitude improvement in computing, optics, and AI. Today’s microdisplays and waveguides cannot occlude the real world entirely without a flip down visor, and their digital objects are not as realistic as what an Apple Vision Pro can produce. There may be no way to create general purpose AR glasses that are stylish, comfortable, and can be used all day long. However, it is certainly possible to build glasses that can manage specific use cases. Qualcomm has a Snapdragon AR1 chipset and reference design available today that spreads the computing – and heat – around the eyewear. In 2019 North launched Focals, AR glasses designed primarily for notifications. North was about to launch its second generation when Google bought the company in 2020 and killed the product line before it got a chance to improve and grow. North also had a unique solution to the problem of interface control – it used a ring with a tiny joystick for your thumb. Wearable Devices has a gesture-based solution to the input dilemma that doesn’t require Apple’s extensive cameras and computing resources, instead using a wrist band (or band for your smartwatch) and AI to interpret nerve response in your hands. Xreal, Lenovo, TCL, and others have display-only glasses on the market today that tether to your phone that could serve as a base for more. If you start with limited AR use cases today and create a product people like, that could be a way to compete not with where Apple is today, but where it is going. It will require that optics and mobile computing capabilities increase faster than Apple can make its pass-through goggles useful out and about.
It is also theoretically possible that other wearable devices take over some of these use cases; Humane is certainly trying.
Competitive Landscape: Winners (Beyond ApplE)
Qualcomm: Meta gets all the attention for investing tens of billions of dollars in VR and renaming the company after its efforts, but Meta’s silicon provider, Qualcomm, has the most to gain if Apple’s entry ignites interest in VR as a category. Qualcomm has invested early and often in silicon for XR and AR, and while it hasn’t had much competition, the market has not taken off to the extent that it would have liked. Apple will never buy Qualcomm chips for the Vision Pro, but Apple’s entry should catalyze market investment with Qualcomm as a primary beneficiary.
TSMC: Any time there’s increased demand for the newest process silicon, TSMC wins. The Apple Vision Pro is built around Apple M2 and R1 chips, which are fabricated by TSMC. Future versions of the Vision line will undoubtedly use more advanced Apple M chips, but, regardless, they’re coming from TSMC, too. All that Qualcomm silicon going to competitors? TSMC.
Microsoft: Microsoft’s HoloLens had all the promise in the world. Like the Apple Vision Pro, Microsoft’s original HoloLens had incredible technical capabilities for the time, and with the expected improvements over time, it could have become a core enterprise tool and then made its way down to consumers. Instead, there was no HoloLens 3.0. Stymied by personnel and technical challenges, Microsoft abandoned mixed reality hardware. However, while the ROI on HoloLens might make its accountants weep, Microsoft knew from the start that Apple was going to enter the market and deliberately separated its software and cloud APIs from its hardware. Microsoft executives told me at the time that its strategy was to profit from software and Azure, no matter what hardware ended up winning. Microsoft is an Apple Vision Pro launch partner for productivity software and should be expected to port Dynamics and enterprise AR apps over as well.
Conclusion: Will the Apple Vision Pro Succeed?
In the short term, Apple will sell every Vision Pro it can make to software developers and tech enthusiasts eager to get on the ground floor of a new computing platform that works within Apple’s ecosystem. Longer term success depends on two factors: what apps get built for it, and how quickly Apple can iterate and get the weight down and comfort up. I’m less concerned about the price; if there are compelling use cases, people will be willing to spend to get those experiences, and then prices will naturally come down over time as the supply chain ramps up to meet demand. It is also worth noting that there are all sorts of enterprise use cases where the Apple Vision Pro is both far more capable and actually less expensive than existing solutions. Apple mostly showed off consumer and personal productivity use cases at WWDC, but the company has made steady gains in the corporate world with the iPhone and Mac; CIOs looking to solve business problems will be among the first to test what VisionOS can do.
Apple has put conditions in place for long-term success, but what that success will look like is a bit of a mystery. That’s just how new computing platforms work: everyone knew that constant connectivity and information access would be hallmarks of smartphones, but nobody predicted that shared narcissism and unlicensed taxis would be killer apps. I can’t wait to use Apple Vision Pro as a virtual monitor system and to watch movies in a virtual theater, but its real value has yet to be coded.
To discuss the implications of this report on your business, product, or investment strategies, contact Techsponential at avi@techsponential.com.