The Vision Pro's real killer app isn't productivity — it's memory

What happened

Christian Selig — the developer behind Apollo for Reddit and, more recently, a string of small utility apps — published a short blog post arguing that the coolest thing he's done with an Apple Vision Pro has nothing to do with productivity, entertainment, or any of the use cases Apple has spent two years marketing. He used it to record spatial videos of his family home before it was sold.

Selig walks through the setup with characteristic understatement: he wore the headset around the house, capturing rooms, corners, the view out of a specific window his mother used to sit next to. When he plays the footage back, he isn't watching a video of the house — he is, in the practical sensory sense the Vision Pro is uniquely good at delivering, standing in it. Scale is preserved. Depth is preserved. The ceiling is where the ceiling was.

"It's the first time a piece of consumer technology has let me record a place the way a camera records a face," he writes, and that framing is doing a lot of work. Photos flatten space. Video flattens space and adds time. Spatial video, captured on a device most people can't afford and viewed on a device even fewer can, is the first medium in the consumer stack that treats a room as the subject rather than the backdrop.

Why it matters

The Vision Pro discourse has been stuck in a loop for two years. Reviewers put it on, try to use it as a laptop replacement, note that it is heavy and expensive and lonely, and hand down a verdict somewhere between "impressive tech demo" and "solution in search of a problem." The device's defenders counter with immersive movies and Mac virtual displays. Neither camp is wrong, exactly, and neither is describing the thing Selig is describing.

What he's describing is a new medium, and new media almost never announce themselves through the use cases their creators marketed. The camcorder was sold on home movies of birthday parties; it ended up defining the aesthetics of an entire generation of independent film. The iPhone was sold on "a phone, an iPod, and an internet communicator"; it ended up rewiring photography, taxis, and attention itself. Apple has spent two years pushing the Vision Pro as a productivity and entertainment device. Selig is quietly pointing out that the actual thing it does — record and replay physical space with fidelity — has no obvious analog in prior consumer hardware, and that this might be where the story is.

The comments on Hacker News are worth reading in full, but the pattern is striking: dozens of engineers, most of them Vision Pro skeptics on the productivity axis, describing their own versions of Selig's use case. Grandparents' houses. Childhood bedrooms. A garage workshop before a parent moved into assisted living. A dorm room the week of graduation. The commercial pitch keeps missing, and the emotional pitch keeps landing without anyone at Apple seeming to notice.

There's a harder question underneath this, which is what happens to the medium when the capture device gets cheaper. Spatial video on the iPhone 15 Pro and later is a shadow of what the Vision Pro records, but it's a real shadow, and Meta's headsets are converging on the playback side from the other direction. If the interesting artifact of spatial computing turns out to be recorded space rather than augmented space, the winners look very different from the roadmap Apple, Meta, and every AR startup has been running on.

What this means for your stack

If you're building on visionOS or any spatial platform, Selig's post is a useful gut-check on where the actual pull is coming from. Most developer effort in the space has gone into ports — productivity apps, media viewers, dashboards floating in the corner of a room. That work is not wasted, but it is competing with the Mac virtual display, which is free, better, and shipped by Apple. The greenfield is on the capture-and-replay side, and it is startlingly empty.

Concretely: there is no serious consumer app for organizing, tagging, or sharing spatial video. There is no equivalent of Photos.app that treats a captured room as a first-class object rather than a video file. There is no export path that lets someone who doesn't own a $3,500 headset experience even a degraded version of the recording. There is no service that treats the capture as an archival object worth backing up on the same terms as a wedding video. Any one of these is a plausible product; all of them together is a category.

There's also a data model problem worth chewing on. Spatial video is fundamentally different from 2D video in ways that current tooling doesn't handle well — you can't scrub through a room the way you scrub through a timeline, and you can't crop it, and thumbnail generation is a nontrivial question. The infrastructure layer here — codecs, storage patterns, streaming primitives for stereoscopic HEVC — is being figured out largely inside Apple, and there's real room for open tooling to move faster than the platform vendor is willing to.

Looking ahead

The Vision Pro will probably not be the device that mainstreams any of this. It's too expensive and too heavy, and Apple's rumored cheaper follow-up is still at least a year out. But Selig's post is a marker for something that's going to matter regardless of which headset wins: the first generation of humans is now capturing physical spaces the way earlier generations captured faces and voices, and in ten years there will be a lot of people who want to walk through a room that no longer exists. The developers who figure out the archive, playback, and sharing layer for that will be building something people actually keep paying for — which is more than can be said for most of what's shipped on visionOS so far.