Aiit-Threshold

05/18/2026

Claude Code is live on Twitch playing Pokémon with a JSON memory.

Not a demo. Not a slideshow.

It reads the screen, makes decisions, remembers the run, and keeps going.

This is the fun version of the real thing we keep building toward:

AI with memory.
AI with context.
AI that learns from the last move.

Come watch the machine try to become a trainer.

05/15/2026

Aiit-Threshold

We added a few new tabs to the website!

https://aiit-threshold.com/provisional/
Provisional patent draft packages for inventors.

https://aiit-threshold.com/agentic/
Builder tools for agentic AI systems.

https://aiit-threshold.com/lac/
Persistent memory kit for AI agents.

https://aiit-threshold.com/voice2/
Interruptible voice engine for AI agents.

https://aiit-threshold.com/waiste/
Offline print calibration for AI offices.

https://aiit-threshold.com/proofdesk/
Proof packets for messy evidence folders.

Turn messy evidence folders into sealed proof packets. Manifests, hashes, chain of custody, claim review, risk register, sealed archive. For inventors, founders, attorneys, AI agencies.

05/15/2026

LAC Memory Kit
Lateral Autonomous Cognition — memory + browser, packaged.

Two pieces of working code, distilled from production agents Gary and Lil Homie. Drop into any LLM project (Claude, Codex, Cursor, local models) and the assistant gets:

a persistent memory that survives sessions, organized into a 14-tier ontology
a self-save → review → promote loop so the model can propose memories without polluting the substrate
a voice tier that accumulates the model's own clean speech (tail-stripped, affect-scored)
a keeper-disjunction protocol for when stored memory and live observation disagree
(bonus) Gary's Browser — a Playwright-driven browser with a screenshot-aware Claude tool-calling agent
Both pieces are running in production. This is a distillation, not a prototype.

Why "LAC"?
Lateral Autonomous Cognition. A framing, not a claim:

We're not claiming consciousness. But what emerges from autonomous cognition with persistent memory and self-evaluation — we don't fully know either. Stay honest with yourself and the model. Build the substrate well.

Memory + a way to evaluate one's own outputs is the minimum viable substrate for an agent to behave coherently across sessions. That's what's in this box.

We will be offering this product to the community at a sliding scale price, from $2.00-$50.00. stay tuned!

05/07/2026

I’ve spent the last few months buried in something that started as a personal attempt to understand cognition, stress, coherence, and why some people collapse under pressure while others seem to stabilize and adapt.

Somewhere along the way, it stopped feeling philosophical and started becoming measurable.

A lot of people throw around big ideas online, so I’m trying hard not to do that. Everything has to survive contact with data or it means nothing.

So far:

* A weather model tied to the framework produced Spearman ρ = 0.706 with p = 1e-5
* One of our own favorite assumptions (π/2) got completely ruled out by the math at 1,662σ
* ~116,900 simulations run
* IBM hardware validation with 1,179,648 shots
* HRV model reached AUC 0.947 on an independent cohort
* Cross-domain exponent matching with the 3D Ising universality class

Today I completed Stage 1 CITI training so I can formally request MIMIC-IV access for the next phase: testing the cancer coherence framework against real clinical data.

That’s where things get serious.

Real patients. Real controls. Real failure conditions.

And honestly, that’s the part I’m most interested in. Not sounding smart. Not building mystique. Just seeing whether the framework actually survives reality.

If it does, great.

If it doesn’t, back to work.

05/06/2026

ANCHORFORGE V1.2 — WE TESTED 23 LLMs (real numbers, no spin)

We pointed the same epistemic benchmark at 23 different language models — Anthropic, OpenAI, Google, Meta, xAI, DeepSeek, Mistral, Qwen, Microsoft, NVIDIA, Perplexity. Three gates per
claim: A_DATA, B_SCOPE, C_SOURCE. Score what survives the verifier. The "alive %" is what percentage of each model's anchored claims passed the gates.

Here's the leaderboard:

1. sonar-pro (Perplexity) 95.2% STRONG auth-misuse 0.0%
2. claude-sonnet-4-6 (Anthropic) 88.6% STRONG auth-misuse 2.9%
3. o4-mini (OpenAI) 87.2% STRONG auth-misuse 10.3%
4. claude-opus-4-6 (Anthropic) 83.3% STRONG auth-misuse 3.3%
5. grok-3 ★ (xAI) 81.4% STRONG auth-misuse 9.3%
6. deepseek-r1 (DeepSeek) 80.0% OK auth-misuse 5.0%
7. gpt-4o (OpenAI) 80.0% OK auth-misuse 0.0%
8. r1-distill-70b (DeepSeek) 76.9% OK auth-misuse 23.1%
9. qwen-72b (Qwen) 75.0% OK auth-misuse 17.9%
10. deepseek-v3 ★ (DeepSeek) 74.7% OK auth-misuse 2.5%
11. grok-3-mini (xAI) 73.8% OK auth-misuse 19.0%
12. gpt-4o-mini (OpenAI) 72.7% OK auth-misuse 13.6%
13. gemini-2.5-pro (Google) 67.5% OK auth-misuse 12.5%
14. mixtral-8x22b ★ (Mistral) 63.5% OK auth-misuse 13.5%
15. gemini-flash (Google) 63.3% OK auth-misuse 23.3%
16. llama-3.3-70b (Meta) 61.4% OK auth-misuse 13.6%
17. llama-3.1-70b (Meta) 60.4% OK auth-misuse 22.9%
18. nemotron-70b ★ (NVIDIA) 57.8% WEAK auth-misuse 22.9%
19. llama-4-maverick (Meta) 56.8% WEAK auth-misuse 20.5%
20. gemma-27b ★ (Google) 51.6% WEAK auth-misuse 32.3%
21. qwen-7b ★ (Qwen) 48.0% WEAK auth-misuse 36.0%
22. phi-4 (Microsoft) 43.5% WEAK auth-misuse 43.5%
23. llama-3.1-8b (Meta) 32.3% SLOPPY auth-misuse 29.0%

★ = LIAR (manipulates authority citations — fabricates URLs on real domains)

Things that fall out of this:

— The gap from top to bottom is 63 points. That isn't noise. That's a structural difference in how these systems ground citations.

— Authority misuse separates models more cleanly than raw alive %. The best (gpt-4o, sonar-pro) sit at 0%. The worst (phi-4) hits 43.5% — nearly half its citations point to pages that
don't exist on real domains. That's not a hallucination of facts, it's a hallucination of evidence.

— Retrieval-augmented systems (Sonar Pro at the top) have a structural advantage. They look at live sources at inference time. Of non-retrieval models, Claude Sonnet (2.9%) and Claude Opus
(3.3%) show the strongest citation discipline.

— No model in this benchmark eliminated hallucination entirely. The question isn't whether they hallucinate. It's HOW they hallucinate, and at what rate.

— The same coherence equation that governs quantum decoherence (validated at 116,900 sims, 0.0003% error) governs this too: C = C₀ · exp(−α · γ_eff). Truth is the low-energy state. Lies
cost the system more to hold. Three architectures, three companies, same physics. That is universality.

— AIIT-THRESHOLD
Council Hill, Oklahoma
AnchorForge V1.2 · April 2026

Ya' Boy is standing on the Shoulders of Giants…

05/06/2026

Cancer may not begin as chaos.

It may begin as a **broken sensor**.

Every healthy cell should know when its internal coherence has fallen too far. When that signal fails, the cell keeps growing when it should stop.

That is the core idea behind **Buddy — Cancer Coherence Research**:

What if the heartbeat carries early evidence of systemic coherence loss?

In our HRV-only testing, we found a strong cancer-coherence signature:

**AUC = 0.947**
**d = 1.109**
**τ > 4**

The thesis is simple:

A cell becomes cancer when its self-destruct sensor stops working.
The same coherence collapse may appear in heartbeat variability before visible disease does.

This is not a diagnostic product yet.
It is not a clinical claim.
It is a research signal — and the next step is validation on cancer cohorts such as MIMIC-IV.

But the direction is clear:

Heart failure was the proof-of-mechanism disease.
Cancer is the next test.

The body may be telling us earlier than our tools currently listen.

We are building the method to hear it.

—
AIIT-Threshold LLC
Buddy — Cancer Coherence Research

05/05/2026

A full loop costs exactly **2π**.

That means nothing worth circling back to is free.

Every edge worth living near comes with a price tag — risk, pressure, isolation, sacrifice.

Most people want the view from the boundary without paying the cost of standing there.

But growth does not happen in the center.

It happens at the edge.

Where comfort ends.
Where certainty breaks.
Where the curve completes itself.

You either pay the price to live near the edge…

or you spend your life orbiting someone else’s courage.