Naming & Figure & GroundAI: Summary
This wide-ranging session, held in the hours before Apple‘s Worldwide Developer Conference, circled around a single underlying question: what is the right relationship between a person, their knowledge, and a designed space — physical or in XR. Conversation moved fluidly between flow states and active engagement, the perceptual default of the human visual system, contested histories and multiple narratives, immersive environments as persuasive “atmospheres,” and the craft of guiding rather than overwhelming a reader. A live demonstration of marked-text “sculpting” in Author grounded the more philosophical threads, while recurring touchstones — the Vatican, Rome, Kashmir, and Stewart Brand‘s How Buildings Learn — kept returning the group to how spaces shape understanding.
Active engagement, not passive consumption, is what moves knowledge into a mind: reading and underlining sit at the bottom of the engagement ladder, while rewriting sits near the top, which makes the prevalence of copy-and-paste a genuine concern for where the brain is actually working.
There is a productive tension between flow and deliberation. Flow suits artistic and generative work, but intellectual work needs a mechanism to step back and engage the prefrontal cortex — so the goal is not to maximize flow but to let a person choose where in the process to sit.
A novel suggestion surfaced that flow state might be measurable, non-intrusively, by analyzing a person’s chat logs with an AI — using the conversation itself as evidence of sustained focus, sidestepping the Heisenberg-like problem that measuring flow tends to break it.
The perceptual “default” is not blank or zero but nature — the visual system evolved expecting the detail of leaves and trees, which is why flat lines and empty surfaces actually demand more processing. An early military video-conferencing system that defaulted to the outline of a person (optimized for crisis-speed transmission) became an example of how “default” is always a design choice, not a neutral absence.
Physical movement is cognitively load-bearing: because the hippocampus binds both spatial and abstract memory, walking through a space — rather than dragging objects with a cursor — genuinely helps thinking, which is the real basis of the memory palace and a strong argument for embodied XR over flat manipulation.
The community still lacks a vocabulary for the basic unit of a thinking space — node, element, knowledge object, concept, person — and that naming gap reflects a deeper unsettledness about what these things actually are. A related realization: people may deserve to be represented as people (avatars, portraits) rather than collapsed into documents, though this is useful for history and distracting for dense academic literature.
Once you step a layer above text, the open question is whether non-text visuals can convey meaning at all, or only direction — a relationship like “there’s something between Doug Engelbart and Ted Nelson” is easy to show spatially, but no visual beats the sentence for the actual content.
History is not “a history.” The strongest framing of the session was the idea of contested or multi-narrative spaces — walking American history as an indigenous person, a woman, or a former slave, or feeling the “ghosts” of Chinese, Pakistani, and Kashmiri perspectives standing around you — so that the environment itself signals “this is not your only path,” nudging the mind to explore what it is missing.
Immersive environments are ancient, not new: the Vatican‘s galleries (heaven across the ceilings, maps of the earth along the walls) were read as a 16th–17th century attempt at a VR experience, persuasion through total atmosphere rather than argument.
Spaces embody a kind of intelligence we rarely design for deliberately. Situated and land artists — James Turrell, Spiral Jetty, Robert Irwin — make the place itself the work, suggesting that interfaces (even 2D ones) could use space far more actively than as a flat surface to place things on.
Flexibility is the cardinal virtue, and it is at risk. How Buildings Learn warns that buildings (and by extension XR spaces) fail when frozen as monuments to their makers; because XR environments are relatively hard to build, there is a real danger of mapping out a single rigid path that everyone is forced into, rather than a “map of infinite variety.” The repeated plea was to keep it fluid, flexible, and adaptable — while still providing enough guidance to avoid the aimless wandering of Second Life.
Sculpting by deletion emerged as a method, demonstrated live: marking the single key sentence in each paragraph (and questioning why a paragraph exists if nothing in it is worth marking) yields an author’s “gourmet selection” that can be read as a standalone overview. The insight that this distilled layer needs to be sharable and encoded — so a reader can choose to read at that level — connects directly to the aims of Origami Text, and raises the deeper question of what one nests for oneself versus for publication.
All rhetoric, and all art, persuades — logically, emotionally, or by titillation — and the group recognized it tends to default to academic, logical critique while neglecting “art level” discussion. The corollary for knowledge systems: sometimes a person wants a “cartoon Socrates,” sometimes a richly textured world, and being honest about that range (rather than wanting it all at once) is itself difficult.
The process-versus-product distinction, drawn from a critique of Hollywood and Mandalorian and Grogu: only the final artifact reaches the audience, just as only the ad — never the marketing plan behind it — is ever seen, which means an interface must carry its meaning in what the user encounters, not in the team’s conceptual model. Simplicity can be a feature, not a failure of ambition.
Skill scales with control: more expressive power demands more effort, like the difference between a plane, a bus, and a bicycle — so good interaction design must offer layers of how engaged a user wishes to be, exactly the scaffolding-versus-freedom balance at the nub of the craft.