What happens to architecture when AI agents and robots can act on behalf of people? Proxy Society investigates how buildings and public spaces can support delegated participation while preserving understandable authority, equitable access and meaningful human presence.
THESIS / 001
CENTRAL PROPOSITION
A proxy can reduce unnecessary travel, waiting or administrative friction. But buildings should not quietly become optimized for automated transactions at the expense of people who arrive in person. The research treats absence and presence as design conditions that must remain compatible.
Q01How should a building distinguish a human visitor from an authorized artificial representative?
Q02Which spatial boundaries should an agent be able to cross, and how should that authority be made legible?
Q03When does delegated participation improve access, and when does it hollow out meaningful human presence?
Q04How can buildings support absence without making physical presence a privilege?
Q05Who absorbs the congestion, verification work, surveillance and maintenance created by automated convenience?
Q06Which spaces should deliberately resist proxy activity in order to preserve care, quiet, ecology, play or intimacy?
03 RESEARCH DIRECTIONS
The project is not about decorating AI with architecture. It asks what spatial decisions change when delegated agency enters buildings, institutions and public life.
When participation can be delegated, architecture must answer a new question: what is still worth being physically present for? The aim is not to privilege presence, but to make absence possible without degrading the right to arrive, stay and participate in person.
Entrances, lifts, service corridors, lockers and thresholds can become explicit interfaces between human authority and machine action. Robot infrastructure is treated as a design hypothesis to test, not a predetermined solution.
Places already carry collective intentions such as public access, clean water, accessibility, safety and ecological protection. Proxy Society asks how those intentions remain operative when automated systems increasingly negotiate access, movement and resources.
PHD PILOT / CIVIC BUILDING
A university library or community centre becomes a living laboratory. A person delegates a bounded task such as reserving a room, collecting an item or communicating an access need. The architectural question is where that authority begins, narrows, becomes visible and returns to a human.
Human intent and human presence arrive together. Staff infer identity, authority and exceptions from the encounter.
The building must distinguish who delegated the task, what is permitted, what expires, and when a human handoff becomes necessary.
Human, proxy or service actor reaches the building edge.
The place checks purpose and authority without requiring excessive surveillance.
Permissions become more specific as the actor moves deeper into shared space.
Objects, information or responsibility transfer at an explicit spatial boundary.
The system returns authorship to a person when consequence, ambiguity or conflict increases.
METHOD / RESEARCH BY DESIGN
The test is simple: can the research identify a spatial design decision that changes the outcome? If the answer is only a better notification or interface, it is not yet architectural research.
Document how people, staff, deliveries, reservations and exceptions already move through a real building.
Translate authority, waiting, handoff, access and conflict into plans, sections and time-based spatial diagrams.
Produce alternative threshold and circulation arrangements rather than only changing the software layer.
Build full-scale or clearly staged spatial prototypes that make delegated authority visible and testable.
Test AI delegation against ordinary booking, fixed automation and human assistance to isolate what architecture actually changes.
Measure access, errors, staff intervention, movement conflicts, comprehension, dignity and perceived control.
ARCHITECTURAL CONTRIBUTION
The project should produce more than a framework for agent ethics. It should leave behind spatial evidence, tested configurations and architectural knowledge that can inform real buildings.
A tested vocabulary for proxy entrances, shared thresholds, transfer zones, human-priority areas and machine circulation.
Evidence-based guidance for when architecture should separate, mix or refuse delegated activity.
Short observational films that capture waiting, misunderstanding, interruption, assistance and informal encounter as spatial phenomena.
1:1 or situated experiments that let participants experience alternative arrangements before deployment.
A framework for deciding when no architectural intervention is preferable to adding new technical infrastructure.
RESEARCH TRAJECTORY
AI helps a person decide what to buy, save and invest. The question begins as interaction design.
AI stops only advising and begins acting. The problem becomes delegation, authority and social meaning.
Delegated action enters shared buildings, streets and institutions. Authority acquires a physical boundary.
EXISTING PROXY SOCIETY MATERIAL
Swimmable Cities treats clean water and embodied public access as an urban condition. WiFi.ee treats connectivity as legible public infrastructure. Human Intent extends delegated agency into buildings, ecological mandates and machine-readable places.
Move from discovering swimming places to asking what makes a waterfront genuinely accessible, healthy and worth inhabiting.
OPEN CASE →Move from describing infrastructure to studying where people can actually sit, connect, get help and remain without friction.
OPEN CASE →Move from agent permissions to physical thresholds, human-priority zones and places that carry persistent public mandates.
OPEN THESIS →ARCHITECTURE RESEARCH POSITION
Proxy Society treats architecture as a governance medium for delegated agency. The research asks how access, authority and handoff become spatially legible, while protecting the ability of people to participate directly in the places that matter.
DISCUSS THE RESEARCHRETURN TO PROXY SOCIETY