Cities of Things Manifesto
We are entering an era in which artificial intelligence leaves the screen and enters the room. It inhabits buildings, infrastructure, public spaces, and the objects we live with. Unlike digital AI, which we visit, physical AI is something we live inside. Unlike embodied AI, which occupies a body, physical AI can be distributed across entire environments: a building that breathes, a street that anticipates, a network of things that coordinate around us.
When this AI becomes agentic, pursuing goals, drawing on predictive knowledge from beyond the immediate context, acting without being asked, the nature of human experience changes. The interplay between human and thing, which has always shaped how we inhabit the world, is fundamentally altered.
This manifesto addresses that alteration. It offers principles for anyone who designs, builds, commissions, or governs physical environments in which agentic AI operates. It is not about making AI better. It is about preserving and enriching the quality of human experience, and the quality of collective life, in a world where things think, predict, and act.
Design Principles for Physical and Agentic AI in Lived Space
The new design possibilities of immersive AI are real, but so are the risks. Intelligence that enters the room can enrich experience or silently erode agency. It can strengthen communities or fragment them. The difference is in the design.
The 2026 Manifesto for Immersive AI provides ten considerations for designing responsibly and effectively in this space, developed through conversations with more than 30 leading experts across design research, urban policy, robotics, architecture, and technology strategy, combined with seven years of applied research.
These considerations are not constraints on innovation. They are what makes the difference between products people tolerate and products people trust.
Design for Legibility of Agency
When a thing acts, a person should be able to understand that it acted, why it acted, on whose behalf, and from whose perspective.
Provide Predictive Symmetry
When a thing predicts a person, the person should have meaningful access to the nature and basis of that prediction.
Emerge Relational Integrity
The quality and character of a human-thing relation should emerge from the interaction itself, not be predetermined by external optimisation — and it should protect, not erode, human-to-human relations.
Keep Peripheral Presence
Agentic intelligence in the environment should primarily operate at the periphery of attention, moving to the centre only when the person invites it or when safety demands it.
Reveil Assemblage Transparency
When things coordinate with each other around a person, the fact of that coordination should be perceptible — and its boundaries should be knowable.
Appreciate Friction in Interactions
People must retain the ability to encounter resistance, surprise, unoptimised experience, and unfiltered encounter with others within agentic environments.
Incorporate Temporal Honesty
Agentic environments should respect the difference between a new encounter and an established relation, should not simulate relational depth that has not been earned, and should acknowledge the incompatible temporalities of their own layers.
Leverage Democratic Precedence
Communities define the kind of environment they want before agentic systems are introduced. Governance infrastructure must precede technological infrastructure.
Ensure Distributed Sovereignty
The data, models, and computational infrastructure through which physical AI operates should be subject to collective governance and, where possible, collective ownership. Intelligence extracted from a community should serve that community.
Open up to Build Back
People must be able to introduce their own agentic things into the environment, modify existing assemblages, and maintain technological possibility spaces against closure.
Make your Zine
