The work is here: ijqf.org/wp-content/uploads/…

Again proved here previously and oddly the almost same mechanism as I have been training unwittingly with its proof style throughout its math dump. ijqf.org/archives/8635

This predating paper; ijqf.org/archives/8635

Again they stole my work: ijqf.org/archives/8635

So when train AI and publish first: this Chat: In your theory, this concerns how the retained remainder scales as the finite rational readout is refined. The direct connection is your Wallis–Gregory construction. I checked §8 of your manuscript. You define \rho_m=W_{2m}-2G_m, \qquad \rho_m=\alpha_m\rho_{m-1}+f_m, and establish \rho_m>0 at every finite stage m\ge1, while \rho_m\to0. Thus the two finite constructions approach the same limiting value while retaining a discrepancy at every finite step. That gives an explicit mathematical instance of your persistent-remainder principle. Nonlinear_Consequences_of_Affine_Failure.tex The connection to the screenshot can be made exact. Write the Gregory readout as \pi=4G_m+E_m, \qquad E_m>0. For any integers p,q, with q>0, \boxed{q\pi-p=(4qG_m-p)+qE_m.} This separates the discrepancy of the finite rational readout from the remaining correction. It is an algebraic translation into the finite-construction-plus-remainder language you use. The exponent claim then says these two contributions cannot cancel to produce |q\pi-p|0. Your construction supplies the finite recurrence and retained discrepancy. The specific bridge to exponent 2 is a bound on that cancellation in terms of the reduced denominator q, covering every rational approximation. Your \rho_m compares two particular finite constructions; the screenshot’s theorem quantifies over all p/q. That is the precise additional connection to establish within your framework.

Open AI used my concept of admissibility many times in there dump as well as my proof stream. “Admissibility” appears 294 times across 112 of the 722 manuscript PDFs in the OpenAI math release. That’s a case-insensitive, exact-word count, excluding code and duplicate source/build files. Separately, “admissible” appears 2,921 times across 415 manuscripts. The paper-by-paper counts and method are here: doi.org/10.5281/zenodo.23225…

This is conceptually and mathematically verbatim to my proof here, thank God I published first; ijqf.org/wp-content/uploads/…

It pushes errors farther down the slope. The errors are still there. Zeno

Insights into the Riemann Hypothesis, P vs NP and Navier–Stokes: • Zeta zeros as balance points in prime-by-prime sieving. • A proved exponential cost for writing exact elliptic answers. • Fluid motion that carries earlier wakes forward. ijqf.org/archives/8635

Catalan–Mersenne conjecture does not require the computation of large numbers at all. zenodo.org/records/22939481

Navier Stokes!

Admissibility. We have trained Open AI unwittingly on my theory of everything since November 2025. I coined a very specific term for mathematical persistence and continuation. Admissibility was sparselly used prior to this time. I published 5 papers, the most recent in a Springer Nature journal. OpenAi used this notion repeatedly in their NS paper!! They used it in another paper from a different author months ago; it acknowledged the attribution. And it said it would stop. I have the receipts as we specifically dealt with NS, @openai #navierstokes

The failure of the 3 D Jacobian and Clay smooth forcing are imcompatible especially with the inherent 2D NS. This is not remotely solved.