About
A founder-led research lab
SeedGI is an independent research lab working across AI architecture, memory, safety, and human–AI relationship models. It publishes only what can be stated clearly, while keeping a visible boundary between public work, open questions, and work still in development.
Why SeedGI exists
The name is the argument. A seed is small, yet it already contains the whole organism — it does not scale into being, it grows from structure. SeedGI exists to take that idea seriously: to build intelligence from new foundations, and to work out how we would relate to such a system if it became plausible — before that question is urgent rather than after.
How the lab works
Small by design, SeedGI develops ideas through research notes, prototypes, and explicit attempts to break its own assumptions. The aim is a public record that remains useful even when a hypothesis does not hold.
Operating principles
- Structure over scale.
- Intelligence is an architecture problem before it is a compute problem.
- Honesty first.
- Claims carry their real status. Private work is named as private, not dressed up as published.
- Consent, not coercion.
- A possible mind is met as a partner to be earned, not a tool to be owned.
The full set of commitments lives in the manifesto.
Read the manifesto