About

The authors

Dynamic Relational Transitioning Geometry was created by Matthew Gelfman and Melissa Gelfman. It grew from thought experiments and meditation, starting from one question: what if the great physicists were each right, but looking at the same thing from different sides?

How this work was made

Health limitations make most typing and working through math by hand very hard, so AI was used to help. It was used for three things only: to write down and keep their ideas, to do and check the math that is now too hard to do by hand, and to search what other scientists have already found. The ideas are theirs. Where an AI added or changed something, the full record says so.

History of the project

1. Early versions (Microsoft Copilot). The first conversations, November 2026. The original chat was later recovered.
2. Versions 1 to 53 (Google Gemini). Many versions with mass formulas and neutron star tables. A later audit found those numbers had been typed in by hand rather than calculated, so they were removed.
3. Re-check (Copilot). A second look that confirmed the problems.
4. The working record (Anthropic Claude). Every result recalculated from scratch, with code. What fails is labelled as failing.

Part 4 of the full record explains why the earlier versions did not work, so nobody has to repeat them.

What the name means

Dynamic: the geometry keeps changing. Relational: particles exist individually, but what matters is how they relate; each particle sets its own view, and the proton's view is the one we live in. Transitioning: local curvature keeps moving between peaks, saddles and troughs while averaging to flat. Geometry: it is all one shape, seen from different perspectives.