Audience
Technologists who are not familiar with Augmented Reality (AR).
Purpose
Convince the audience that AR is the next paradigm of computing.
What is AR?
Augmented Reality (AR) is technology that can overlay digital information onto the real world. There are multiple ways to achieve this—either by actually projecting onto an object (like a spotlight) or by virtually projecting (like a heads up display). The former approach is not scalable—we’d need different projectors for every object in the world. We will talk about the latter approach, which is achieved via a digital device (smartphone, glasses).
To head off confusion, AR in this document also allows for virtual interaction between real and non-real objects (e.g. virtual makeup sticking to your face in AR even as you move around).
Types of AR and Benefits
Like most digital devices, there are 2 dimensions for AR devices. The first is input: can the device capture information about its surroundings? Does it have a camera to “see” or a mic to “hear”? The second is output: does the device relay information to the user by voice or by vision? Google Glass had visual and audio inputs as well as visual and audio outputs. Most AR devices today only have audio inputs and outputs.
Output on AR devices improves the transmission speed of information. It also allows for better information comprehension. Tangibly, you won’t need to take your phone out everytime you need something. But even further than that, the device can proactively provide information. This could range from the mundane (“This restaurant has your favorite dish”) to the life-saving (“You should check out that mole on your arm”).
Input is the game-changer though. Without input, an AR device is no more than a smartphone on your face. But with input, it can become your guardian, your scribe, your nutritionist, your personal trainer, and so much more. An AR device with input can show you proper form in tennis. It can identify poisonous mushrooms. It may even warn you of others’ erratic driving. Whatever information passes through your eyes and ears, an AR device with input could filter and enhance. Input-enabled AR devices will be a complete upgrade to your senses.
Output on AR devices improves the transmission speed of information. It also allows for better information comprehension. Tangibly, you won’t need to take your phone out everytime you need something. But even further than that, the device can proactively provide information. This could range from the mundane (“This restaurant has your favorite dish”) to the life-saving (“You should check out that mole on your arm”).
So, not only will AR 10x our senses but it will also help us with tasks we didn’t even remember or know of.
Downsides to AR
People care about privacy. First, people don’t want others recording them without consent. Second, people don’t want anyone else (read: corporations) seeing their private data. If I wore AR glasses, you wouldn’t want me to discreetly record you. Likewise, I don’t want the AR company, from whom I bought the glasses, to have unchecked access to what I see.
Devices can be hacked. If I can hack your device’s inputs or outputs, I control what you see. I can add in objects that aren’t there or mask objects that are. The safeguard here is that AR devices don’t replace what you see, they only overlay. So if you turn off the device, no harm is done.
Some may fear putting electronics so close to their face for extended periods of time. Galaxy Note 7s instilled a fear of Lithium-ion batteries that people may not easily lose. Then again, people put their phones in vulnerable positions all the time.
Advertising may proliferate on AR platforms. To the ad providers (e.g. Google), every inch of the world is now available as advertising space with zero marginal cost.
Glasses are the primary form for AR devices. To some, this may be a deterrent—glasses are not universally attractive. However, the majority of Americans do use glasses or contacts, so there is still a big market there.
Will users adopt AR? Historical guides
Historical examples help us answer this question. The biggest consumer hardware launches in recent times have been the iPhone and the Oculus series. The iPhone is one of the most ubiquitous devices the world over and it truly launched the smartphone revolution. The Oculus, though a great device, has shipped <100M. AR is more like the iPhone than an Oculus headset. Similar factors that propelled the iPhone will propel AR headsets.
The crucial difference between the iPhone and Oculus is that the iPhone is of value wherever you are, for almost any use case. This is because it had a web connection. You could check and respond to your email anywhere. You could navigate directions on the fly. You could message people instantly or even video call them (in later iPhone iterations). Simply put, the iPhone constantly helped you navigate reality.
The Oculus, however, is its own world. That’s great for games and some escapism. However, the Oculus and VR can only minimally help you in actual reality. If you’re trying to navigate to a restaurant, the Oculus is of no help. And it will be a long time before virtual worlds are detailed enough to supplant real ones.
The distinction can be summed as: the Oculus helps you escape everyday life while the iPhone is everyday life. AR headsets will start off as enhanced iPhones, so it too will be your everyday life. The headset will be with you throughout the day and can help you accomplish any task. AR will be a painkiller, not a vitamin.
Will users adopt AR? Key factors
Style matters. Ugly consumer devices are not common. A popular AR headset must be stylish, especially since helpfulness won’t be a guarantee in the beginning. The iPhone was stylish. Google Glass was not.
Price matters. It can’t be too expensive since users will naturally ask “why not just buy a new smartphone?” Plus, a small userbase means little content developed for the device. It also can’t be too cheap or consumers will think it’s of low quality. An initial hypothesis is the $500-800 range makes sense, comparable to top-end gaming consoles and VR headsets and middle-range smartphones.
Controls matter. Users will want to control whether the camera (or other inputs) is on, just like their smart home devices. Users will want to control what data is shared with the headset manufacturer, especially if they wear the device at home or in personal moments.
Above all, experience matters. Of course, if there’s nothing to do on the device, then no one will want to buy it. But lots of mediocre content is also not a seller. What helped the iPhone was that its native web browsing and email were exponentially better than competitors. Similarly, AR devices will need to nail the fundamentals: navigation, email, messaging, and video. If the UX on these is at least as good as a smartphone’s, then the convenience of the AR device will carry the rest of the way. This is just the beginning though. The real power is in the device as a platform for others to create on. Incentivize developers to create awesome “apps” and experiences, and people will buy and keep coming back. The chicken-and-egg problem, of attracting users to attract developers to attract users, can be solved by focusing on the fundamentals, just as the iPhone did. That generated the hype to start the cycle.
Timing
People have visualized AR as far back as 1901. It suddenly seemed possible back in 2011 when Google Glass appeared. But they didn’t take off and AR has since been relegated to face filters and furniture shopping. The next 3 years, however, are the right time to invest in AR.
Large tech companies are working^1 on AR glasses. Meta itself has 10,000 people in its AR/VR division^2. Almost every AR hardware company has promised glasses within the next 2 years. This surplus of competition will lead to lots of innovation and differentiation in the next few years; many types of AR glasses will come out—mostly not interoperable with each other (a big opportunity by itself). More importantly, as big tech goes, so does the industry.
The hardware needed for powerful, useful AR glasses has matured. Waveguide technology allows you to see the real world and a projected image together: you can overlay virtual objects onto the real world. Battery life is now upwards of 3 hours as well. Significantly, the performance of chips like the M1 mean that the AR glasses themselves can handle processing without a phone tether—and in a compact, weightless form. A user won’t need to carry around their phone just to use their AR glasses. A small company (<250 employees) called Nreal recently released AR glasses that are close to consumer-ready. They were lightweight, able to project images onto the real world, were not unstylish, and cost $600. All this points to one conclusion: it’s no longer hardware stopping AR adoption.
Users are tired of incremental phone upgrades. Pre-launch, only 10% of iPhone users planned to upgrade to the iPhone 13^3. Every few years, the iPhone wows (like the iPhone X), but in between it’s all incremental^4. Users want that big upgrade, that life-changing device. As we’ve talked about, AR glasses promise to be that life-changing device. But users are ready for that new device now. They weren’t in 2013 with Google Glass: iPhone was less than 6 years old. Smart phones are ubiquitous and not “hype” anymore; you’re not a tech influencer or early adopter just by owning the latest iPhone. But you would be with AR glasses.
However, economic development may hinder AR glass development. The main factor is a potential recession. It’s beyond the scope of this memo, but Wall Street thinks a recession may be coming^5. In recessions, consumer spending decreases, especially on luxury electronics^6. An old iPhone and a pair of AR glasses may not seem so different when the price difference may be that months’ groceries. But, it’s not a given that the recession will arrive or that it will even stay for long. Even if the recession arrives, the short-term target audience for AR glasses will be early adopters—techies with ample spending money. Mass appeal is still further out.
The chip shortage may hinder companies’ AR efforts. Why use valuable chips on unproven AR glasses instead of on the proven cash cows of laptops and iPhones? It’s a classic case of the Innovator’s Dilemma. But there are also signs that the chip shortage is easing, especially for chips like the M1^7. In the end, AR glasses won’t be mass adopted for a few years, by which point the chip shortage will no longer be a problem.
In conclusion, 2022-2023 promises to shower us with new AR glasses. The threat of recession and chip shortages loom too. But if just one company can nail style and user experience, 2023 will be the year we go from mobile to glasses.
Startups and AR
So, AR glasses are coming in 2023. If you’re interested in founding an AR company, what should you do?
Not hardware. Hardware is really expensive. And time-consuming. And requires lots of research. These constraints make it suited for big tech. Google can afford to throw thousands of people and hundreds of millions of dollars
AR’s effects on technology, platforms, FAANG
The insane effort and cost needed to build good hardware makes it likely that a big tech company will be the one to do it. Just like Apple with the iPhone, Google via acquisition of Android, and Meta via acquisition of Oculus. It will be an all-out war among the tech giants because hardware is the most immutable part of a user’s tech stack. A user can make the decision to cancel their Netflix account or to switch from Safari to Chrome any day. But users really only think of switching their hardware once a year (e.g. when new phones come out).
Hardware is also the foundation of the stack; the company that wins the hardware layer wins no matter what. Apple doesn’t care that Google Chrome is more prevalent than Safari; users use Chrome on Apple devices. In fact, Google pays Apple $10+ billion every year to make Google Search the default on iPhones. The company that wins the hardware layer gets to dictate what is built on top of it. Other companies’ software has to fit the dictates of the hardware company. That’s partly why Apple commands 30% of revenue on the App Store: because it won the hardware. All this to say, tech giants will view the AR device battle as existential. If Google dominates, it could ruin Apple’s future. If Facebook wins, it could build its own search engine on top of the hardware layer. Among Google, Apple, and Facebook, the losing companies may not survive the next 20 years.
In all likelihood, we will see a duopoly just as we do now. There will be the luxury AR glasses—the new status symbols. And there will be the cheaper AR glasses used by everyone else. This seems bound to happen because no one product can appeal to both high-income and low-income markets. Even monopolies have different product lines for different incomes. Apple seems the most likely to win the high-income market. Not only do they already have that luxury reputation and ecosystem lock-in, they also have a knack for UX and they have the powerful M1 chips. Once Apple comes out with their glasses, the rest of the world will look to them for inspiration. Only then will we see the “Android” of AR glasses released for other users. Developers, looking to code for as few platforms as possible, will naturally look to the dominant companies, leading to a duopoly again.
The glasses industry will be affected too. Rayban (and Luxottica, the glasses monopoly) can’t directly compete with smart glasses. But, they still have valuable knowledge (style) and user brand loyalty. So these companies will likely partner with tech giants (like Rayban did with Meta) to offer their frames as a base for the AR components. Apple may even just buy some brands out and vertically integrate, as they do with most other products. One thing we may see is a reduction in styles—style won’t be the differentiator between early AR glasses products. Comfort and helpfulness will be more important instead.
AR will birth the next decacorns. Unicorns pop up after technological revolutions. Amazon, Google, PayPal, Facebook, etc. were only possible because of the web. Stripe, Square, Airbnb, Uber, ByteDance, Instacart, Doordash were possible because of mobile. The same can be said of AI as well. As we mentioned, the potential for AR is unknown. We can’t predict what kinds of companies will be built. But suffice to say, a large user base and tons of input data (visual and aural) will lead to extraordinary innovation and legendary companies built by daring founders in the next 2 decades.
What does this mean for you, the aspiring startup founder?
Now is an excellent time to build for AR: powerful AR glasses are just around the corner. However, you should not enter hardware now. That would be entering the lion’s den and you likely will not have the money, the talent, or the time to stand a chance. Let the tech giants build the hardware.
The best bet is software.
Figuring out new UX and UI experiences (e.g. what’s the “cursor” or “swipe to scroll” for AR headsets?) will be crucial in the next few years. Startups in that space will likely get acquired by the tech giants, leading to big paydays and to making outsized impacts on the industry.
Interoperability will be a mess for a while as every tech giant tries to create their own glasses. Creating single APIs to develop for any glasses at once will be a godsend to developers. There may not be many winners in this space but the winner will win big. And it can only be a startup that does this kind of work.
Developers will be key to making this work. Tech giants will focus on incentivizing developers too but there are always more APIs to create. For example, an easy API for computer vision model creation will be absolutely necessary for AR as many pieces of software may want to interact with different visual objects and in many different ways (e.g. app A wants to identify every storefront whereas app B wants to identify every moving car). Additionally, easy testing and deployment software will be necessary too.
The biggest bet will be in content creation. With AR glasses, a person’s vision becomes their canvas. Imagine TikTok on AR glasses—“day in my life” videos may take over the world. Similarly, we will need just such content creation mechanisms in AR. Low-code and no-code are the key paradigms here.
There is so much more. Taking literally any company and saying “but how would we do this in AR?” represents a legitimate opportunity for disruption.
<!-- Footnotes themselves at the bottom. -->