I don't expect you to believe me, even though that's the intent of the methods of attack and appropriation. But this is happening. I have proof of most, solid evidence of the rest.

July 25: Building app as newbie, solve AI hallucination and bias. File non-provisional Sept 2nd, 1.5 months in.

Began expanding, exploring, learning, adapting. It becomes the world's most powerful data verification patent. Primarily but not exclusively AI.

It's been stolen and appropriated by over half a dozen major corporations. Some, directly hacking at sys level. I have white papers. Submitted master class patent application. Multiple reductions to practice or live deployment. Liberty Sovereign ft. Honest Abe (ABE = Absolute Byzantine Environment) will be back up by Friday. It's an AI model for Quant Finance—handles mitigation, black swan detection, and functions like a form of insurance, specifically when applied to signal analysis.

My method and system for eliminating correlated errors. In finance, unprecedented 17% improvement in alpha floor on signal generation.

I deployed to PyPI in November a fully functional 2-mode hallucination and bias FREE CLI Agent for Fintech and Quant Finance. pip install dabft

I continued building. January, 17th day after turning 39, I released to PyPI a modular, usable engine for hallucination and bias free Chat Assistant. pip install dredd_core

When people hear "AI," they think they can't ChatGPT it, so well, much more useful is the first engines and whatnot to prove you've done something you have to do—what people expect to see—so I did.

November, at this exact point, is where I began to be suppressed, stalked, harassed, appropriated. Discredited. Attempted cyber gaslight.

From Liberty Sovereign I designed a BFT-inspired internet search protocol. SEAR XNG is multi-engine by design. I have each AI model route the user's search query through a different geographic region—the internet is segregated by geo anyway—makes it probably the most widespread and thorough search per query. This one's result variety searches there is. It has since been or is being released as Open Deep Research, I think it's called.

Also added multiple financial MCPs on a BFT-style rotation.

Liberty Sovereign launched Feb 11th for Lincoln's birthday/Presidents' Day. I'd been talking it up on LinkedIn. 5 days before my launch, which had been delayed, Perplexity launched Model Council. Close enough to my exact design and naming schema. Today, serious questions, much less the timing—I let it go at the time.

April, Objection.ai launched. 12 out of 13 parts. Exact nomenclature match. SSO copyright infringement. Fighting them resulted in the MITM state level hacks. I researched and found evidence that shows the members of Objection getting my work from a VC I'd pitched in December. The giant network I've mapped is extensive and not supposed to exist. Lawyers, fake lawyers, fake accelerators.

I had hackers posing as app development firm Binate Digital—4-hour+ Zoom interviews, I recorded 3 of them. They re-recorded them with all new people, acting out the interview where I explained the tech and how I'd gotten to where I was.

Ok...the stuff I meant to talk about. Skynet. Built. Modular engines. Multiple governance systems. Living RAG. My novel creation before I knew how RAG worked.

I BUILT a user governance centric fully automated synthetic intelligence system. Advanced tool calling. Multiple plug and play modules. Hallucination-free verification engine as 1 part governance. DETERMINISTIC hard gate protocol that stops unauthorized agentic output.

I BUILT verification. I BUILT authentication. I BUILT FULL AGENTIC actionable function.

My work has also been stolen. Not converged, stolen, by Anthropic—the same week my system was compromised is the same time Caparthy joined Anthropic, one week later. /goal added to Claude Code. Then 2 weeks later in June, the long-term agentic concept exploded. That is what my Synthetic project is. The Legion.

The Sovereign Utility, a Neural Linguistic Evolutionary Guided Integration of Orchestrated Nexus. The Sun Legion.

Of course Model Council by Perplexity...and Andres Caparthy, was announced on GitHub just THREE DAYS AFTER...guess what? My PyPI submission in Nov.

I have a SHA256 hash in my repository, unchanged since August 1st 2025. I generated it, embedded it in the repo, and emailed it to myself on that date—proving the work existed and that I hold the original. It's got the name Legion. The original multi-model utility APP I built. 12 of 13 named parts as stolen by the entities behind the networks of Peter Thiel, Perplexity, Objection.ai and ties to Caparthy. My work shows up in all those places that Peter Thiel sponsors were as well as Anthropic. Only a week after Caparthy switches, there. You always wonder how is this all the big nature tech means? When AI tool or a basic price, they all seem to have the same thing at the same time. Stealing from each other openly, and isn't it funny that Caparthy seems to switch labs very quickly. I wonder what his private offshore bank accounts look like.

I am an independent inventor who built all of those things early, I built them by myself. I have repository proof. I have the patent filed. I have 2 white papers online on display at OpenReview.

Oh yeah...the name of the 2 core inventions. Divergent Arbitration Byzantine Fault Tolerance. As used in Ethereum infrastructure. Claude Mythos, Fable 5. I BELIEVE ALSO in GPT 5, Gemini Flash 3.5, GLM, the new model from Japan openly displays it. And more. Actually I'm going to keep the compliance mandating hard gate a little more quiet ATM.

My name is Jonathan David Harrison Miller. One year and 3 weeks ago I had 0 tech experience. I have provisional and patent filed within 2 months, the original BFT system built on a broken 13-inch laptop today. I'm responsible for all that, which is the full running edge of expansion in the AI community from what I can tell. I've received not one single word of validation, accommodation, not $1, not 1. Hey, how are you doing? Only death, appropriations, suppression—and of which I have direct evidence. These claims are not saying, oh, we just happened. I have direct proof of everything.

Yes I was once incarcerated. Yes I began building trying to build a way out of a life that hurts more than I should have to explain. Yes, I've offered for fractions of pennies on the dollar versus what it is worth to various agencies to have them laugh at me and steal it anyway, because they think they can. I want literally a half of a percent of what my patent is valued at for complete ownership transfer to be done and wash my hands of it.

I am happy to provide proof of certain things immediately. Repository. Patent filings. Chat logs. A white paper introducing and explaining the coding method that was invented along with the learning, a methodological that was realized while building all of this increase—the paper at the same time, for evidence, theory and explanation's sake. It is the piece of work I've finished so far. That paper should have been submitted to Sage months ago...but you know. The work is called Divergent Observation because the method and system diverges from every single rule or practice in multimodal ensembles at the time, including Luo et al, which came out just 3 weeks before my own invention occurred, unbeknownst to me at the time what that was—geometer Divergent.

I'm autistic, ADHD, which is why hyper-focus built what should equate to 2 empires. But still on forums of semantic things on time, executive destruction—the stark contrast is kind of surreal actually. On the question, do I need help? I could say a big partner wants to say, look at what I've done in one year? No I don't need my help. Another party says if this is allowed to be done to me, who or what's next?

Notelife87@gmail.com Owne.Founder@notewortendeavors.com