For months, one of the biggest questions inside our AI Factory has been deceptively simple: How do we stop treating B-roll as a bolt-on step and start treating visual production as an intelligent, inspectable system? We now have an answer. OpenMontage has completed its acceptance path and is now the default media engine in our AI Factory. This was not a “we installed it and it launched” milestone. We pushed the system through real production tests, deliberately difficult visual requirements, failed generations, licensing checks, approval gates, mixed-media fallbacks, rendering problems, timing bugs, and independent validation. By the time the final Organic Orchard proof passed, OpenMontage had produced a valid 90-second H.264 video at 1920×1080 with AAC audio, passed all six approved visual slots, preserved the original narration and SRT unchanged, and cleared 31 out of 31 tests. The Old Problem: B-Roll Was Too Narrow Traditional automated video workflows often follow a patte...

Bottom Line: We are not replacing our AI Factory with Hermes, and we are not planning to depend on Nous-hosted inference for normal Factory operation. We are testing a local-first supervisor stack: Hermes Agent as the orchestration layer, Ollama as the local inference server, and Qwen as the local reasoning model . Hermes decides what should happen next while deterministic Factory gates decide whether the work is actually valid, approved, and finished. This article documents our current build-in-public architecture decision while the Deep Dive AI Factory is still actively integrating OpenMontage. The Question Changed From “What Can We Automate?” to “Who Runs the Factory?” For months, our local Deep Dive AI Factory has grown one solved problem at a time. We built a YouTube production pipeline. Then review gates. Then metadata safeguards. Then Blogger and Facebook publishing bridges. Then visual QA, local models, ComfyUI, a Factory Command Center, startup health checks, and a gro...

Bottom Line: Building an AI system is surprisingly easy compared with cleaning one up. The hard part is not deleting old code. It is deciding which lessons, safeguards, and proven workflows deserve to survive. This sentence is the publication verification marker for the AI Factory cleanup article. When Experimentation Works a Little Too Well Give an AI agent enough time, Git access, a few APIs, some Python, and a mildly obsessive human operator, and eventually you do not have a project anymore. You have an archaeological site. There are working tools. Half-working tools. Experiments that became production systems without anyone officially declaring them production systems. Old branches containing one brilliant idea buried under hundreds of outdated files. Documentation describing architectures that stopped being true months ago. New systems that solved problems the old systems were specifically designed to solve. Then someone asks the terrifyingly simple question: “Can we cle...

Michigan Baseball • 30th Season 30 Years of Luglife! Baseball, Big Lug, Honolulu Blue and three decades of Lansing memories at Jackson Field. Bottom line: Thirty years after the Lansing Lugnuts first took the field, the formula still works: summer weather, minor-league baseball, a ridiculous mascot, a giveaway you suddenly cannot live without, and another memory added to three decades of Luglife. Thirty years is a long time for anything to survive, especially something built around summer nights, foul balls, mascot chaos, cheap beer and the stubborn belief that a minor-league baseball game can still feel like a major event. That is basically the story of the Lansing Lugnuts. Tonight, sitting above the field at Jackson Field while the grounds crew finished its work and the late-August sun stretched across the infield, the anniversary suddenly felt less like a number printed on a logo and more like a piece of Lansing history that is still alive. ...

Deep Dive AI Blogger Publishing Skill Test This is a controlled test of the Deep Dive AI Blogger publishing workflow. The purpose of this post is to confirm that ChatGPT can hand a validated Blogger post to the Deep Dive Blogger Queue, the local Windows bridge can publish it through the existing Blogger API runtime, and the workflow can return a verified live result. What This Test Confirms The Deep Dive Blogger queue is reachable. The Blogger publishing bridge can consume a standard publish job. The post can be published through the established Deep Dive workflow rather than manual copy and paste. Live-page verification can confirm that the expected body content appears after publication. Test Content Only This is not production editorial content. It is intentionally short and exists only to verify the publishing pipeline on August 29, 2026. Verification marker: DD-BLOGGER-TEST-2026-08-29-1240-DETROIT

I’m Trying to Make ChatGPT Do the Boring Part: From Blog Idea to Blogger to Facebook Bottom Line I’ve been working on a simple idea that sounds almost too obvious: Give ChatGPT a topic, have it write the blog post, publish that post to Blogger, then automatically turn around and share it on Facebook. That’s it. No complicated marketing funnel. No 47-step automation diagram. No dashboard that looks like NASA mission control. Just: Topic → ChatGPT → Blogger → Facebook The funny part is that writing the article is probably the easiest piece. The real work is getting all three systems to reliably hand the job off to one another without me becoming the unpaid copy-and-paste department. And that is exactly what I’ve been trying to fix. The Problem Isn’t Writing Anymore AI has already changed the hardest part of blogging. A few years ago, writing a decent 1,500-word article could eat up an evening. Now I can give ChatGPT a subject, some context, a few rules abou...

MRI Safety Navigator: Building a Better Workflow for MRI Implant Verification For decades, MRI technologists have dealt with the same basic problem. A patient arrives for an MRI examination with an implanted device. Sometimes there is an implant card. Sometimes there is a model number. Sometimes there is an operative note. Sometimes there is little more than a manufacturer name, a partial device description, or a patient who remembers being told that the implant was “MRI compatible.” Then the real work begins. The MRI technologist may have to search manufacturer websites, MRI safety databases, technical manuals, device records, implant documentation, and other sources before determining whether the examination can proceed and under what conditions. The information often exists. The problem is that the information is fragmented. That is the problem MRI Safety Navigator is being designed to solve. This Is Not Another MRI Implant List There are already excellen...

How to Use MRI Safety Navigator: A Practical Implant Research Workflow for MRI Technologists There is a familiar kind of MRI delay that starts with one short sentence: “The patient has an implant.” Sometimes the implant card is complete. Sometimes the operative note is useful. Sometimes the chart says only “stent,” “pacemaker,” “shunt,” or “stimulator,” and the MRI team has to turn that fragment into an exact device identity, a complete system, a current manufacturer document, a scanner-specific set of conditions, and a defensible local decision. MRI Safety Navigator was built to make that research process more structured. Open MRI Safety Navigator and start an implant research case → MRI Safety Navigator is designed to organize implant research, not replace the people responsible for final MRI authorization. It is not an automatic clearance engine. Current manufacturer labeling, the complete implanted system, the exact planned examination, facility policy, and appropriate...

MRI Safety Navigator: A Better Way to Research Implants Every experienced MRI technologist knows the moment. The schedule is moving, the patient is waiting, and an implant suddenly becomes the most important object in the building. Sometimes there is a clean implant card with a manufacturer and exact model. Sometimes there is a partial operative note, an old device description, or a patient who understandably remembers only that “they put something in.” Then the real work begins: identify the complete system, locate the controlling documentation, compare every condition with the planned examination, and document why the scan can—or cannot—move forward. After roughly 25 years working around MRI safety decisions, I wanted a better starting point. That led to the MRI Safety Navigator , a free browser-based research and documentation tool designed to organize the first stages of implant investigation. Bottom Line MRI Safety Navigator helps a technologist search for a device using...