The Future of AR and “The Next Big Thing” Illusion
We have a natural tendency to assume that technology is a one-way street, that each new technology invalidates everything that came before it. First analog, then digital, then AI, and so forth.
But the reality is that every technology is simply a solution to a set of very particular problems that people face.
So they only invalidate other technologies to the extent that they solve these problems better than those other technologies.
Let’s take cars as an example.
It’s been 140 years since the car was invented.
Yet we still have trains. We still have trams. We still have horses. We still have bikes. And we still have walking.
Because the reason why cars are everywhere has nothing to do with them once being considered “the technology of the future” or “the next big thing”. Those words aren’t some magical spell to pop a technology into existence.
Cars are everywhere because they’re simply good at quickly transporting a small number of people > 1km and < 200km.
Not < 1km. Not > 200km. And not a large number of people. Other technologies are still better at handling those cases.
They were better than horses at doing this one thing, so people started using cars instead.
And we just see lots of cars because many people want to transport a small number of people > 1km and < 200km every day.
Things are that simple.
I say this because society keeps forgetting this every time a new technology comes out.
New technologies are always “the next big thing” that will “change everything”, and everyone ignores the fact that these are just solution to problems.
If they do a better job solving problems than other technologies, they survive. If they don’t, they die.
And, for those that survive, if the problems they solve are important for a lot of people, they’ll become influential technologies. If the problems aren’t that important, they don’t.
So there’s nothing magical about technology.
To go through another example, let’s look at cryptocurrency.
At its core, cryptocurrency is simply a way to exchange value without some centralized entity messing with your ability to do that. That’s the purpose it was designed with.
So why is anyone surprised when, 17 years after its invention, it does exactly that?
It didn’t disrupt existing currencies in the western world, because those governments generally don’t mess with the currency enough for most people to care.
In other words, inflation has been low and the average person isn’t meaningfully affected by currency controls or de-banking.
But cryptocurrency has seen at least some degree of adoption in places like Venezuela, where the government does make exchange meaningfully harder by messing with things.
So, in places like this, cryptocurrency provides some advantages over existing technologies for solving that problem, namely using the black market to deal in US dollars.
And, more importantly, cryptocurrency did massively disrupt a financially meaningful sector: crime.
It did so because exchanging value without the government messing with you is basically the most important problem that criminals face, and using a cryptocurrency is much better than having to launder money through 400 pizzerias.
In other words, cryptocurrency has grown to a respectable degree not because it’s “the technology of the future”, but because it solved a particular set of problems better than the alternatives.
It didn’t and won’t invalidate government-controlled currency anytime soon, because control by even a moderately competent centralized entity offers inherent advantages over a cryptocurrency, like lower volatility and lower transaction costs.
But it’s easy to make these claims in hindsight. So let’s test whether things are actually this simple by making a prediction about the future.
Here’s the prediction: augmented reality glasses are currently purely early adopter products, but I predict they’ll become a billion-user product by 2032.
Why? Because they solve problems that lots of people have and they’ll do so better than the alternatives.
And, importantly, there are no major technical barriers in the way of them doing that.
To explain why, let’s go through some of the other technologies that are currently used to interact with digital apps:
Smartphones
This technology is fundamentally just a small device with a two-dimensional screen that you control by touching it.
Because it’s small and requires no peripherals, it’s incredibly convenient.
This means smartphones are a fantastic solution for doing something quickly, whether it’s taking a photo or sending a quick text or playing a game while commuting.
But those same properties make it bad at other things.
For one, it isn’t and will never be the best way to do something precise or complicated. There’s just a fundamental limit to how much you can control things by tapping on a small screen with your fingers.
This will inevitably make it arduous to fill in a spreadsheet, create a CAD model, play a game that has more than two buttons, or do anything else that requires complicated input.
Other than input, the fact that smartphones are small means that they fundamentally can’t have as much processing power as a bigger thing. So any task that does require serious processing is better achieved through bigger things.
This is why smartphones didn’t miraculously invalidate personal computers, despite once being “the technology of the future” and “the next big thing”.
Understanding these fundamental limitations is what made companies like Supercell so great.
While everyone else was trying to make computer games on mobile, they made genuine mobile games, with simple controls and gameplay loops designed for short sessions (like a commute). So they went on to change the industry.
Personal computers
A personal computer is a big two-dimensional screen, which you control with a mouse and keyboard (that allow for very precise and complicated input).
Is it super convenient? No.
While smartphones exist, nobody will pull out their laptop to quickly send a text or check the weather or play some game for 5 minutes on a crowded subway.
But personal computers are still the best way to do things that rely on processing power and complicated input, like playing videogames with in-depth controls and impressive graphics.
Do they provide the most processing power of any technology? No. That’s supercomputers and cloud computing.
But the extra inconvenience and price you’d have to pay for more processing power isn’t worth it for anyone who isn’t doing drug discovery or weather prediction.
Virtual reality
Virtual reality, by comparison, is a three-dimensional screen that you control with hand motions and a few buttons.
The three-dimensional screen requires you to put a bulky thing on your head. This makes the technology, by definition, insanely inconvenient.
It’s the price for the “reality” part of “virtual reality”, and the same inconvenience will apply to the high fidelity brain-computer interfaces of the future.
So, while some people might still use a laptop on the subway, no one will ever use a virtual reality headset in those contexts.
Or, for that matter, in any other context that doesn’t justify a huge amount of effort.
Because, if you try using a headset, you’ll quickly realize that you’ll be taking that bulky thing off and putting it back on every time you want to drink water, eat something, talk to a friend or colleague, check your phone, go to the bathroom, or do anything in the real world.
Apart from being wildly inconvenient, it’s also a technology that makes anyone stand out very negatively in any public space. So it’ll never be used outside of your living room either.
The reason is that headset is ugly or, more scientifically, does nothing to emphasize the physical features that make us attractive to others. It just hides our entire face, which is usually the most important factor behind attraction.
And it also blocks our eyes, which is problem because we look at people’s eyes to interact with them and determine what they want to do. When we can’t do that, we implicitly feel that people are distant or weird and suspicious.
Nothing will change this, which is why Apple wasted so much time trying to solve the problem with their goofy projected eyes on the Vision Pro (and we saw it as goofy precisely because we’re so good at reading eyes, so we instantly notice when things are off).
However, that same bulky three-dimensional screen that comes with all of these downsides also makes it the best way to feel like you’re actually a part of a virtual world.
And feeling like you’re actually part of a virtual world really matters for some use cases:
It’s much better and cheaper to train a nuclear engineer in a virtual reality simulation of their real environment, rather than expecting them to learn procedures from a textbook or spending a lot of resources to construct a sufficiently realistic training environment in real life.
Virtual reality can be a much better substitute for face-to-face interaction than texting or video calling, which matters when you’re in a long-distance relationship or have severe social anxiety (or there’s a pandemic).
And many games are all about getting lost into a virtual world. And it's much easier to get lost in that world when you can actually look around it and even touch it (although this use case is still bottlenecked by issues, like how VR games still make you feel motion sickness).
Then there’s the controls part. Virtual reality lets you control things with your three-dimensional hand motions.
So suddenly you can make things feel visceral and physically involved. It’s you swinging that sword or doing the dance move, rather than a character on some screen.
This basically makes it a version of the Wii or Kinect that allows for much more complicated and precise input, but at the cost of convenience.
There’s enough demand for this that some games have already become serious contenders against computer and mobile games in their genres, like Beat Saber (a dancing game), Blade and Sorcery (a swordfighting game), and Gorilla Tag (a tag game).
However, controlling things with three-dimensional hand motions and a few buttons also means that it’s just very inefficient for navigating any kind of interface quickly.
Tangibly: instead of moving a 10 gram finger a few centimeters to get to another place in the interface (with a mouse or screen tap), you have to move a 4 kilogram arm 50 centimeters to do the same.
It’s hard to overstate how annoying this gets after a while. And, when combined with the fundamental bulkiness of its three-dimensional screen, it’s why virtual reality isn’t and will never be the best solution for doing everyday tasks.
And, no, using a keyboard doesn’t change things. It makes things better, but you still have to deal with taking the thing off and putting it back on 47 times a day.
In a nutshell, virtual reality is also just good at some things and bad at others.
We had a virtual reality bubble because people didn’t understand what those things were, and Meta pumped the GDP of a small country into use cases that anyone who’s tried a headset knew would fail.
Augmented reality glasses
So, finally, we get to augmented reality glasses.
These are small three-dimensional screens on your face that you control with your hand motions.
Because the screens are small and because there are no peripherals, it’s incredibly convenient.
In fact, it’s arguably even more convenient than mobile, since you don’t even have to spend the tiny amount of effort it takes to pull something out of your pocket.
But, because it’s small and you have to see through the screen for them to function as glasses, it will also inherently be terrible for any task that depends on processing power or impressive graphics.
So it won’t compete with work tasks or the computer games or even mobile games that have any reliance on those things.
However, the fact that you can see through the screen also makes it possible to show three-dimensional things directly in the real world.
This enables showing you stuff in advance or even based on what you’re looking at, like the three-dimensional route to your destination or the exact motion you’re supposed to do for a workout.
And, even more importantly, it can show you things based on what you just did, like playing satisfying animations and sound effects when you complete something successfully.
Together, these seemingly trivial features make it possible to solve countless problems better than any other technology:
Want to see information about something very quickly, like whether this item is cheaper at another store? You no longer need to pull out your phone and navigate 3 different websites to find the answer. It can happen the moment you look at a store item. Sounds mundane, but that’s what consumer technology is.
How about watching videos or reading a chat while lifting weights, instead of purely pulling up TikTok in between sets or having to rely on the gym having a TV (which may or may not show a channel you’re interested in)? Because there are many situations where you can’t rest your phone against anything or where you’re moving around or in an orientation that makes it hard to look at a phone screen. Not a problem for glasses.
Or maybe you completed that equation in your homework? Here’s a progress bar and a level up effect to make you feel like you accomplished something. And here’s a detailed explanation on what you can improve on, delivered instantly instead of 2 weeks later when the assignment gets graded.
This will, in effect, enable the complete gamification of reality. For better or worse.
Every task that you can think of will have the equivalent of a Duolingo, since the bing bing wahoo dopamine hit of a satisfying animation the moment you do something will always outcompete the boring reality of delayed gratification.
The only remaining barrier to this Duolingo dystopia is that, while current AI models are more than up to the task, they’re still too slow.
Even a two second delay is enough to make those gamified rewards feel way less satisfying, to the point of even being annoying if you’re doing lots of tasks quickly.
And having to wait a few seconds for the model to look something up completely negates any convenience advantage the glasses have over a smartphone.
It’s a dealbreaker for the entire technology.
But this problem is more than solvable. It’s way, way easier to know that you’re looking at a $3 can of beans and not a $2 corncob, than it is to operate a car or robot autonomously.
So models just need to get slightly faster. And there’s no way they won’t, when they’ve improved literally 100x in just the past few years. No major innovations are needed beyond what even big, lumbering companies have already proven they can do in very recent history.
Then, moving on from the screens, augmented reality glasses also use hand motions for control. But, unlike virtual reality, there are no buttons this time.
So, without buttons, they’ll be even worse than virtual reality for tasks that require complicated inputs.
Especially because, unlike with virtual reality controllers, mice, or even phone screens, you’re literally just poking at air. There’s no clicking or vibrating sensation or even just the feeling of successfully touching something.
Not only does that feel incredibly unsatisfying, it also makes it very hard to precisely control things. Because it’s hard to know if you, say, moved your finger enough to perform an action (like a click) when there’s nothing to tell your finger that you did.
So you’re left with an implicit feeling of “oh I guess that worked?” every single time you do something.
This, alongside the bulkiness of a virtual reality device, is why Apple’s Vision Pro failed for productivity use cases.
Put simply, poking air seems like cool sci-fi stuff, but it’s a terrible way to navigate complicated interfaces.
But, just like with mobile, there are plenty of tasks that don’t require complicated inputs, like all of the simple gamified examples I mentioned earlier.
And it’s possible that someone will eventually invent better control methods, but I think the technology will reach the mainstream before then.
Then we have the last major feature of augmented reality glasses: the fact that they’re on your face, always looking at what you’re looking at.
This will make privacy a serious problem even for the average joe who, let’s face it, hasn’t cared at all about their smartphone being the most effective spying device in history.
The reason is that augmented reality glasses make it very, very easy to violate someone else’s privacy.
Here’s why: recording someone with a smartphone is inherently noticeable.
You have to point the camera at someone to record them, which forces you to hold the smartphone at either an angle or a height that’s clearly different from how you’d normally hold it.
So, as long as they don’t see someone using their smartphone like that, people will generally feel safe, knowing that some creep isn’t recording them.
And would-be creeps will implicitly feel pressure to not do this, because they know there’s a pretty high risk that they get caught and embarrassed.
The problem with augmented reality glasses is that there’s no way to tell if someone is recording you.
The glasses are automatically at eye height and therefore will be able to record you, even if the person is seemingly staring through a window that’s way to your side.
And people aren’t stupid. Everyone will realize this, especially when we’ll start seeing videos of creeps subtly recording women or people doing extremely embarrassing things in settings where nobody would’ve assumed they were being recorded (like the bedroom or hanging out with close friends).
And, yes, you can do things like add a red light that shows if someone is explicitly recording video.
But this is incredibly easy to hack (you just need to disable a single LED) and a lot of the glasses’ functionality relies on constantly recording what you see.
So, red light or not, there will always be a video feed and it will always exist somewhere that opens it up for exploitation — be it by creeps, hackers, or governments.
All that being said, I don’t think this will stop the technology. For the simple reason that bing bing wahoo dopamine hits and even marginally higher convenience are like crack to our monkey brains.
The privacy issues will never be fully prevented but, unlike smartphones, users will genuinely care enough that privacy features and privacy-first brand position will matter for the first time for the average user.
Regardless, when we consider that there’s a huge list of tasks that would benefit from gamification or convenience beyond what smartphones provide, I think this will be a billion-user technology.
Conclusion
People tend to assume that new technologies are fundamentally unpredictable and unknowable phenomena, and that the “next big thing” will always disrupt everything everything that came before it.
But the reality is that new technologies are simply solutions to a well-defined set of problems and they’ll only change to the extent that they solve those problems better than alternatives.
And here’s the controversial part of my argument: I think these problems are so well-defined and human needs are so unchanging that it’s possible to make reasonable predictions about the future of a technology.
Yes, truly accurate prediction is impossible, especially when it comes to timelines.
But I think we can beat a coin flip, as long as we experiment with that technology at an early stage and understand what fundamental trade-offs it makes.
So, to test this theory, here are the predictions I made about the future of augmented reality glasses, which are currently still early adopter products that I haven’t even tested myself (as of 03/03/2026).
Gamified mobile apps, like weightlifting or dieting apps, will be completely disrupted by augmented reality glasses. Because having to manually jot what you did into an app after every task was so annoying that it was the biggest bottleneck for these apps.
It becomes possible to gamify literally everything. This makes the world a Duolingo dystopia, where everyone is constantly doing things just for the sake of points and streaks. But it’ll be better than the current TikTok dystopia, since actually healthy habits like going to the gym or running or even doing homework will suddenly be in the same ballpark of engagement as TikTok.
Podcasts will be impacted, since they’re uniquely good at providing entertainment while you’re doing something like chores. Augmented reality glasses will enable gamified chore apps and the ability to play videos during chores, which will be much more engaging for a lot of people. They’ll still exist because they take up less attention than videos or apps, but they’ll have a smaller role in our daily lives.
Mobile games won’t be significantly disrupted. Because even mobile’s limited input and minimal processing power is way better than what augmented reality glasses can offer. The minor effort that pulling out your phone requires is going to be more than justified by mobile’s superior graphics and controls.
Virtual reality games that rely entirely on physicality will be disrupted, so things like dancing and workout games. Augmented reality glasses will simply be absurdly more convenient (and far less sweaty), while still providing the hand motion-based controls that these games rely on.
Virtual reality games and applications that rely at least somewhat on the feeling of being in a virtual world will be completely untouched. Augmented reality glasses won’t be able to compete against a bulky three-dimensional screen that doesn’t need to be see-through. And nobody wants to escape into their 30 depressing square meter apartment, rather than a huge fantasy world.
The personal computer will be untouched. Because augmented reality glasses are even worse than mobile at all the things that computers are good at.
Privacy issues with augmented reality glasses will be so noticeable that there will be significant backlash and privacy-first brand positioning will genuinely matter for the first time in consumer technology. There will be countless scandals related to government spying and creeps recording women. But the technology will grow regardless, because there will be some OKish solutions to distract people from this risk and because bing bing wahoo dopamine always wins.
The tasks that augmented reality glasses are the best at are so common that they will become a billion-user product by 2032, primarily stealing time away from mobile devices and previously non-digitized aspects of life. And because privacy concerns, AI model latency, form factor, and price are all solvable problems for big companies, even if no real innovation occurs from this point on.
