BioConst
Medical and biological parameters: what changes inside living systems and how to read those changes without overclaiming.
Source-backed topology
Physical AI is a stack, not a model.
PhysDB maps the silicon side of intelligence: robot bodies, world models, policies, data, simulation, control, compute, and safety boundaries.
28 nodes40 edges7 domains28 connected
Batch 1 frame
Carbon, silicon, cosmos.
PhysDB replaces Pockot as the third high-level knowledge-graph venture. Pockot can still return later as a product branch under the silicon map.
PhysDB
Physical AI and robotics: how models meet bodies, sensors, control loops, data, simulation, compute, and evaluation.
CosmosExplore
Aerospace and space exploration: cost structure, physical constraints, mission architecture, and off-Earth infrastructure.
First visual asset
Robots are where models meet matter.
PhysDB starts from the full stack because model-only explanations break down at the point of contact. The first graph is intentionally bounded: source-backed nodes, guarded edges, visible corrections.
Receipts before claims
Knowledge graph first, hype later never.
PhysDB v1 includes a source registry, anti-misinterpretation registry, and correction ledger before it adds breadth.
Primary sources, research papers, and technical docs with scoped use.
Each concept has a caveat and source trail.
Every relationship has a context scope and overread warning.