Why Spatial Computing Still Feels Solo

Why Spatial Computing Still Feels Solo

Spatial computing still feels solo because immersion is often designed for one wearer while access, context, and control remain uneven for everyone else. A mirrored television can show what the wearer sees, but it does not give a nearby person the same place in the experience.

The problem is not simply that a headset covers the eyes. It is that the product creates several participation gaps at once.

One room can contain different realities

flowchart LR
    A["Headset wearer"] --> B["Immersive view, input, sound, and agency"]
    C["Nearby nonwearer"] --> D["Partial screen, delayed context, little agency"]
    E["Remote participant"] --> F["Avatar or stream, different controls"]
    G["Platform creator"] --> H["Rules, identity, moderation, and access"]

The wearer may receive depth, scale, spatial audio, tracked hands, and the ability to act. A person on the couch may see a cropped, delayed, or disconnected view on a flat screen. A remote participant may appear as an avatar. A person passing through the room may not know whether the wearer can see or hear them.

Calling all of those people "users" hides the design problem. They occupy different roles with different information and power.

Immersion can remove shared context

In Venture Step E083, I described using Quest as compelling for solo movie viewing and less convincing as a casual group activity. The problem was not that the screen looked bad. The screen was the part that worked. The problem was that friends nearby could not naturally understand and join what I was seeing.

Casting the headset view to a television helps with visibility. It can still introduce latency, setup, disconnections, and a basic asymmetry: one person controls the world while everyone else watches their camera.

Research has treated that asymmetry as a design problem for years. ShareVR, a 2017 CHI project, combined projection, a tracked handheld display, and physical interaction so a non-headset participant could see and affect the same environment. In a small user study, the researchers reported higher enjoyment, presence, and social interaction than their gamepad-and-television baseline. ShareVR paper

The important lesson is not that every living room needs projectors. It is that a spectator view is weaker than a designed role.

Shared participation has five layers

A system can be multi-user online and still fail several layers. Two headset owners may share a virtual room while excluding the person physically sitting between them. A mixed-reality application may show the real room but still give only the wearer access to controls and meaning.

Bystanders need an interaction channel

A headset changes ordinary social signals. Eye contact disappears. The bystander may not know whether the wearer can hear them. Touching the wearer can be startling, while speaking may be ignored by the application or masked by audio.

A 2020 study of bystander interruption surveyed 100 people and ran a follow-up lab study with 16 participants. It found that the relationship to the wearer influenced comfort and strategy, and that people often combined speech and touch when interrupting someone they knew. Bystander Interruption of VR Users

That is a product requirement, not just etiquette. A spatial system should provide awareness of an approaching person, show the bystander whether an interruption was received, and let the wearer pause without losing progress.

Representation can reduce the gap

A 2024 IEEE ISMAR study examined ways to represent non-headset observers both inside the headset and on a two-dimensional screen. The researchers reported that including the observer improved the shared experience, while preferences differed between wearers and observers. Wearers tended to prefer real-world visual representations. Observers did not agree on one ideal form. Open-access research record

The result resists a simple avatar solution. Presence is not only whether a body appears. It includes whether the representation is understandable, respectful, timely, and useful for the current task.

More realistic worlds do not automatically become more social

At Connect 2025, Meta described Horizon Engine, prompt-assisted Horizon Studio tools, higher concurrency, Hyperscape Capture, and Meta Horizon TV. Those improvements can make worlds easier to create, larger, and more visually convincing. Meta Connect 2025 recap

None of those improvements alone gives a nearby nonwearer agency. A photorealistic room scan can still be a private scene. A virtual cinema can still be one person's screen. A world with 100 online participants can still isolate the headset wearer from the person beside them.

This is why "more social" cannot be measured only by concurrent avatars. The product must account for the physical room, remote network, device mix, identity system, and transition between them.

Cross-device participation should be intentional

A strong spatial experience defines meaningful roles for people on phones, televisions, browsers, speakers, controllers, and no device at all.

The nonwearer should not receive a weak copy of the headset view by default. They may need an overhead map, instructions hidden from the wearer, moderation controls, a camera, voting, audio, or the ability to place an object into the scene. Asymmetry can be useful when each role has a reason to exist.

The system should also expose state. Who is present? What can each person see? Is the microphone active? Is the room being mapped? Is a camera recording? What action would pause or leave? Those answers reduce social uncertainty.

Spatial computing becomes shared when the relationship is designed

Display quality, field of view, tracking, and latency still matter. They determine whether the wearer believes the world. Shared computing adds another standard: whether the people involved can understand one another and act together.

The most useful test is to put one headset in a room with several people and follow every role. Measure the time until the first shared action, not only the time until the wearer enters the world. Observe who explains the interface, who gets bored, who can interrupt, who controls the session, and whether the group wants to repeat it.

Spatial computing will stop feeling solo when the nonwearer is no longer an afterthought. The experience has to include a gradient of participation, from a quick glance to full immersion, without requiring everyone to disappear into matching headsets.

AI assistance was used for research organization, drafting, and validation. Publication remains unauthorized.