AI is moving into the physical world. We’re building what comes next.
Three theses on what comes next.
- Thesis 01 PSI
- Thesis 02 MIR
- Thesis 03 PRACTICE
THESIS 01
Physical Intelligence Goes Personal
WHAT WE OBSERVE
Physical Intelligence is advancing rapidly across foundation models, robotics, 3D, simulation and media.Most of that progress is still being built around machines, specialist workflows and professional applications.
WHAT WE BELIEVE
The next frontier is personal.As these capabilities become cheaper and more accessible, Physical Intelligence will move into everyday life — helping individuals understand, navigate and act on the physical world around them.
BUILD — PERSONAL SPATIAL INTELLIGENCE
We’re building AI for the physical reality of everyday life.That means building beneath the surface too: a spatial foundation that makes a person’s real spaces usable by AI.
THESIS 02
Imagination Has Become Cheap
WHAT WE OBSERVE
AI has dramatically lowered the cost of imagining, exploring and generating new possibilities.But the path from an idea to physical reality remains largely traditional, fragmented and high-friction.
WHAT WE BELIEVE
A new consumer need is emerging.People need a better way to turn what they can now imagine with AI into something grounded enough to become real.
BUILD — MAKE IT REAL
We’re building a personal architect agent for people who want to change a space.Our current prototype grounds a chosen design in the space as it is — turning a picture into the start of a project.
THESIS 03
The Built World Is Still Far from AI-Native
WHAT WE OBSERVE
AI is rapidly changing how knowledge work is done.But the professions that shape the physical world still rely on fragmented workflows, specialist tools and large amounts of manual coordination.
WHAT WE BELIEVE
The next step is not simply giving professionals more AI tools.The practices that shape the physical world need to become AI-native — with intelligence embedded into judgement, knowledge, decisions and workflows from the start.
BUILD — AI-NATIVE PRACTICE
We’re building towards our own AI-native architecture practice, run by a group of talented architects.The test is the whole practice — how AI can automate work and carry intelligence through the loop as opportunities become commissions, commissions become real places, and every project makes the next one better.
The AI-Native Practice
Working model + future direction
- Opportunity
- Commission
- Understand
- Design
- Coordinate & Resolve
- Deliver
- Learn
- Better next project
- new Opportunity
AI automates work and carries intelligence through the practice loop.
- Revenue / FTE
- Labour hours / project
- Throughput
- Cycle time
- Rework
- Margin
- Repeat / referral

FOUNDER
Vito Chen
Architect · Spatial Intelligence Researcher · Founder, ArchMinds
Vito Chen is an architect researching spatial intelligence and the relationship between AI, space and everyday life.
That interest began in civil engineering and architecture: disciplines that teach you to think in physical systems and constraints, and about the people who live with what gets built.
Both are about what happens when an idea meets reality. As AI moves into the physical world, ArchMinds is where Vito researches and builds around that question, starting from everyday life.


