Why Apple's Vision Pro is a brilliant prototype for a product nobody asked for
A year in, the most interesting thing about Apple's headset isn't the hardware. It's what Apple is learning from us wearing it.
A year after launch, the Vision Pro has become something Apple has never really had before: a product on sale, in a store, that Apple itself seems unsure about.
Ask a Genius Bar employee about it and you’ll get the practiced smile they use when a customer brings in a seven-year-old iMac. Ask a developer and you’ll get a shrug. Ask a reviewer and you’ll get a version of the same piece — beautiful, heavy, expensive, isolating, and I can’t quite figure out who it’s for. We’ve written that piece. Three times.
This is our fourth piece, and it’s a different one. Because I’ve come to believe, after six months of daily use and a shelf of competitor headsets next to my desk, that the Vision Pro isn’t a failed product. It’s a running field test. And Apple is the one running it — on us.
The device is not the product
The Vision Pro is a $3,500 headset with two 4K micro-OLED displays, a custom M2/R1 chip combo, twelve cameras, six microphones, and an external battery that tethers to your hip via a single white cable that looks embarrassed to be there.
It’s the best-engineered piece of consumer electronics I have ever put on my face. It is also the loneliest. After six months with it, I have used it more than I expected, and I have used it with other people approximately zero times. There is no version of the Vision Pro that includes a second person. The device knows this. Apple knows this.
And yet Apple keeps shipping it.
The prototype theory
Here is what I think is actually happening.
Apple has roughly 1.4 billion active devices in the wild. iPhones, iPads, Macs, Watches. Every one of them is a sensor for the next one. The App Store tells Apple what apps people use. The Health app tells them what vitals people track. The Photos app tells them what moments people capture.
Then Apple ships a Vision Pro. Twelve cameras. Six mics. Every session logs how long you wore it, which apps you opened, which environments you chose, whether you used passthrough or immersive, how your eyes moved, how your hands moved, where you put windows in space.
Apple is not selling 300,000 Vision Pros because it wants 300,000 customers. It’s buying 300,000 longitudinal studies.
What the data is probably telling them
I don’t know what Apple’s telemetry says. But I know what my own use pattern says, and I know what my friends’ use patterns say, and it’s a near-unanimous shape: early enthusiasm, followed by a quiet collapse into three use cases.
Long flights. I have worn this thing on six flights. It is the best in-flight entertainment device ever made, full stop. A 120-inch screen in the middle seat is a miracle.
Movies at home, alone. When my wife goes to bed early, I watch the second half of whatever we were watching, in a dark room, on a cinema-sized display, with spatial audio. It is a better home theater than any home theater I could build.
Occasional spatial video. Playing back the birthday footage from last August in spatial is genuinely moving. I don’t do it often. When I do, I cry.
That’s it. That’s the whole product.
Apple sees this too. I would bet a significant fraction of my net worth that inside Apple Park there is a chart of median session lengths by use case, and “airplane movies” has a curve that looks like Mauna Loa.
What Apple is building toward
This is where I think the prototype theory gets interesting.
Vision Pro is not a product. It’s a staging ground for the two products Apple actually wants to ship: a glasses-form-factor AR device that weighs under 80 grams and costs under $1,500, and a spatial content ecosystem that’s large enough for that device to land into.
Neither of those exists yet. The hardware isn’t ready. The content isn’t ready. The social protocol isn’t ready. We don’t know, collectively, how to be around each other while wearing eyes-on-your-face computers.
So Apple shipped the heavy version, to the smallest group of people willing to wear one, for the longest time they’d wear it, in the most private spaces they’d wear it in. It learns what we do, what we don’t do, what makes us take it off, what makes us put it back on. When the glasses version ships, Apple will have spent six years of instrumentation time on the problem. Nobody else will have that.
The argument against
The honest argument against this read is that Apple is not that patient anymore. That the Vision Pro isn’t a field test — it’s a genuine miss, and the reason Apple is quiet about it is that Apple doesn’t know what to do next. That the Vision Pro team is being reshuffled, the Vision Pro 2 has been delayed a year, and the entire division is on a watchlist.
All of those things are also true. It’s possible to run a field test that is, financially, a failure.
But the data is still real. The longitudinal studies are still running. Whatever Apple’s internal chaos looks like, the telemetry lake is still filling up. Someone, somewhere in Cupertino, is going to build the next thing on top of what they learned here. That thing won’t be the Vision Pro. It will be better than the Vision Pro. And the Vision Pro will turn out, in retrospect, to have been the expensive, heavy, lonely, brilliant prototype that taught Apple how to make it.
The verdict (mixed, with optimism)
Don’t buy one. Unless you fly a lot and like movies and have the money, in which case, you already did.
But don’t dismiss it either. There is a real, genuine product in here, eight years from now, that your kid is going to wear on a school trip. The thing on Apple Stores right now is the rough draft. The rough draft is good enough that I have kept mine. That is not nothing.
Matt has reviewed the Vision Pro three times and will probably review it again. He has no relationship with Apple’s Vision Pro team. They have not asked for our opinion.
“The Vision Pro is Apple's most expensive developer kit. We paid $3,499 to be beta testers. And that might have been the plan.”