A question to hold, because everybody in the industry is circling it and almost nobody will say it plainly. Who is this for?
Preorders for Apple’s Vision Pro opened on Friday morning and the first allocation went quickly, with delivery dates for later orders sliding into March. The device goes on sale in the United States on 2 February, in Apple’s own stores, starting at three thousand four hundred and ninety nine dollars.
Set the question aside and look at the object, because the object answers more of it than the commentary does.
What the thing is
It is a headset, worn over the eyes, with an external battery on a cable. It contains two internal displays of very high pixel density, a set of cameras and depth sensors pointing outward, cameras pointing inward at the eyes, and enough processing to composite the outside world onto the inside displays in real time.
That last property is the one that matters and the one the category name obscures. This is not a device that shows you a virtual world instead of the room. It shows you the room, through cameras, with things added. The technical term is passthrough, and the entire experience rests on doing it fast enough and accurately enough that your brain accepts the reconstruction as the room.
The interface has no controller. It tracks where the eyes are looking and reads small finger gestures, so selection is a glance and a pinch. That combination is the actual product claim: not that a computer can be worn, which has been demonstrated many times, but that one can be operated without learning anything.
What the price is made of
Three and a half thousand dollars is not a margin decision in the ordinary sense, and it is worth understanding what is inside it before treating it as a statement about positioning.
The displays are the expensive part. Putting a high resolution panel close enough to an eye that individual pixels are invisible requires pixel densities far beyond a phone, on panels that are physically tiny, and the manufacturing yields on those are poor. A low yield means a high unit cost and no easy path down except time.
The sensor array and the real time processing add more. So does the mechanical work: the thing has to sit on a face for an hour without becoming unbearable, which is an engineering problem that consumes materials and iteration.
So the price is largely a bill of materials that has not yet had a decade of volume applied to it. That is a different situation from a product priced high to signal luxury, and it points at a different trajectory.
What the shipping model tells you
The detail most worth attention here is not the price, and it is easy to skip past. It is that this is sold in person, in Apple’s own stores, with a fitting.
The device needs to match a face. Interpupillary distance has to be set, the light seal has to fit the shape of the head, and people who wear glasses need corrective inserts specified. A product that requires measurement before it works is a product that cannot simply be shipped to an address, and Apple has built the launch around that constraint rather than around it.
United States only, at first. That is not caution about demand; it is a statement about how many trained staff and how much fitting capacity exist.
Which tells you the first year is supply shaped rather than demand shaped, and that the numbers reported about preorders, whatever they turn out to be, will be measuring the size of the allocation rather than the size of the appetite.
What it competes with, which is not what people think
The comparison being made everywhere is to other headsets, most of which cost a fraction of this and sell to people who want games.
That is the wrong frame for a device at this price. Three and a half thousand dollars is not competing with a games console. It is competing with a laptop and a large monitor, or with a workstation, and it is being bought by people who will attempt to work in it.
The real question for that buyer is mundane and unglamorous: can you type. Not whether the displays are good, which by every account they are, but whether four hours of actual work is possible, with text, on a surface, with a keyboard, without the neck complaining and without the passthrough latency becoming intolerable.
Nobody outside Apple has run that test at length yet, because the device has not shipped. It is the test that decides whether this is a category or an accessory.
One more property of the object deserves attention because it is the part most likely to decide ordinary use. The battery is external, on a cable, and it runs for a couple of hours.
A device that is tethered to a thing in your pocket and that stops after two hours is not a device you forget you are wearing. Every design decision described above works toward making the interface disappear, and the power architecture works against it, because it introduces a cable, a weight and a clock. That tension is not a flaw somebody missed. It is the honest consequence of the displays and the processing, and it will be resolved by batteries improving rather than by anything Apple can decide.
The answer to the question
So, who is it for.
On the evidence of the object itself, the first version is for developers and for people whose work justifies the price on a spreadsheet. The fitting requirement, the single market, the price, and the absence of a controller all describe a device that is being introduced carefully to a small number of people who will build things for it.
That is a legitimate way to start a platform and it has worked before. It is also the phase at which most platforms die, because the interesting software arrives only if enough developers believe the installed base will grow, and the installed base grows only if the interesting software arrives.
The number to watch over the next eighteen months is therefore not units sold. It is how many applications exist that are worth the weight of the thing on your face, and whether any of them is something you cannot do sitting at a desk with two monitors. If that application exists, this becomes a category. If it does not, this becomes a very good demonstration of displays that will turn up in something else.
