Spatial Intelligence
When intelligence leaves the screen and enters physical space, context becomes continuous. The design problem shifts from displaying information to placing it where attention already is.
Why this matters
Screen-based tools ask a person to leave the environment they are working in, translate what they see into a query, and translate the answer back. In work that happens in physical space, that round trip is where most of the cost sits.
Spatial delivery removes the translation but introduces a harder constraint: attention is finite and already committed. Information placed poorly is worse than information withheld, because it competes with the environment itself.
This makes restraint a technical requirement rather than an aesthetic preference. The question is not how much can be shown, but what earns a place in someone's field of view and for how long.
The OODA perspective
The systems OODA is engineering treat placement, timing, and persistence as core parameters rather than presentation details. What appears, where, and when it disappears is part of the architecture.
Spatial context is also continuous in a way screen sessions are not. There is no natural start and end, so state, memory, and governance have to be designed for an environment that is always partially observed.
Human authority stays in the same position it holds elsewhere in the framework. Spatial systems extend perception and recall; they do not take over the decision or the responsibility for it.
Design & engineering implications
- Implication 01
Budget attention explicitly
Treat field of view as a constrained resource with a declared budget. Every element competing for it should have a stated trigger, a priority, and a condition under which it yields.
- Implication 02
Anchor to the environment, not the viewport
Information tied to a place, object, or task can be reasoned about and dismissed. Information tied to the display follows the person everywhere and becomes noise the moment the task changes.
- Implication 03
Design the quiet state first
The default should be near-empty, with density earned by relevance. Specifying what the system looks like when nothing needs attention is more revealing than specifying its richest view.
- Implication 04
Plan for continuous sessions
Persistence, resumption, and hand-off between people need first-class treatment, because a spatial session does not end cleanly the way a screen session does.
Questions under active investigation
- What signals reliably indicate that a person's attention is available, without inferring more about them than is appropriate to collect?
- How should spatial context be retained between sessions in a way that is useful and still governed?
- Where does spatial delivery genuinely outperform a screen, and where is it simply a more expensive way to show the same thing?
Where this goes next
Spatial work is where augmented intelligence stops being a metaphor and becomes an engineering constraint. The capability areas describe how OODA approaches that build, and a program conversation is where the specific environment defines what is worth placing in view.