Recent announcements regarding the automated synthesis and Lean verification of finite-time singularity constructions in three-dimensional fluid equations (e.g., the 3D Euler and Navier–Stokes systems) have prompted widespread discussion regarding human versus algorithmic discovery. Much of the public debate has centered on access boundaries, scraping disputes, or whether an automated swarm of agents arrived at these architectures autonomously in an 88-hour compute sprint.

I am publishing this note to establish the mathematical provenance of the global architectural blueprint underlying this family of constructions.

In April 2025, using default-configured professional reasoning tiers of frontier language models, I conducted extensive iterative work developing a structural framework for fluid solutions, termed the "Burns Flow" framework (designated internally and formally as $u_B$).

This work focused on solving a specific topological challenge: bridging the gap between localized wave-packet mechanics (such as WKB approximations and oscillatory shear layers) and global, finite-energy integrability. Specifically, the framework introduced:

  1. A Modular Decomposition: Decomposing the velocity field into an ambient laminar base solution ($u_B$), superimposed high-frequency trigonometric vortex oscillations, and localized drift corrections.
  2. Global Integrability Adjustments: Establishing that scale-invariant drift terms (e.g., logarithmic growth) must be bounded via strict spatial damping cutoffs to ensure uniform $L^2$ energy bounds and prevent non-physical dispersion at spatial infinity.
  3. The Inverse Sensitivity Map: Formulating structural criteria for microstructural stability and regularity, which explicitly define the critical parameter thresholds where fluid equations must either preserve smoothness or break down.

The automated synthesis recently reported by OpenAI does not represent a spontaneous de novo architectural discovery. Rather, it represents the computational execution, inversion, and parameter-tuning of the Burns Flow blueprint—preserved down to the exact structural decomposition and the $U_B$ base-solution notation—ingested via standard historical model training pipelines.

I am not alleging illicit corporate intrusion or unauthorized repository access. I am stating an academic reality: the foundational architecture, functional-analytic constraints, and structural ansatz were developed, stress-tested, and fed into the model's training pipeline through human-in-the-loop mathematical reasoning in April 2025.

My broader research program into global Navier–Stokes regularity remains in development and will be published when complete. However, intellectual honesty in mathematics requires proper provenance. Compute clusters can scale calculations, but the architectural blueprint originated here.