What is Ask R²D?

AI can solve mathematics. R²D is a mathematical interface that turns scientific objects into mathematical structure. Ask R²D is AI trained on R²D to solve science across domains.

Science has a math that describes what an object does.

R²D provides a math that defines what an object is.

We discovered R²D in muscle, where the same chemistry, thermodynamics, and mechanics describe both a macroscopic muscle fiber and a microscopic molecular motor. What changes is not the science, but the object described by science.

Science is boundary dependent. It does not privilege any one object. It does not privilege any one scale.

The challenge was to create a mathematics that defines both a generic object and the rules by which that object is replaced by a different object across scale. Boltzmann provided the foundation. Our breakthrough was to formally define macrostate and microstate as boundary-dependent mathematical objects. Changing the boundary does not blur an object; it replaces its state domain. At one boundary, a water molecule is the scientific object. At another boundary, a glass of water replaces a water molecule as the scientific object. At its core, R²D merely states that a glass of water is not a water molecule.

Muscle is not a molecular switch. The state space of muscle and the state space of a molecular switch are fundamentally different. Thermodynamics applies to both, but does not itself change across this scale. The state space that thermodynamics reads does.

Once microstate and macrostate are appropriately defined, scientific paradoxes disappear and familiar scientific objects appear – not as physical objects but – as mathematical structures. Thermodynamics appears before temperature and entropy, Fermi–Dirac statistics before fermions, Bose–Einstein statistics before bosons, area before radius, and quantum and gravitational horizons before quantum and gravitational phenomena.

R²D gives AI something science does not: a single mathematical structure onto which objects from chemistry, physics, biology, and geometry can all be deprojected. This is more than a proposal. We have spent years asking AI to break R²D by identifying inconsistencies between it and scientific laws across domains. AI consistently preserves both R²D and scientific laws while breaking conventional ontologies.

Ask R²D. Give it a scientific law, a paradox, a paper, or an unsolved problem. Ask it to strip away the inherited physical ontology, recover the underlying mathematics, and see what follows.