Untangling the Web

While I’m still fine-tuning the tech for Karellen’s epic Davos debut

—spoiler alert: it’s coming to the Karellen playlist soon

—I couldn’t resist sharing the actual speech early. It’s too good to keep under wraps!

Check out the link, get a sneak peek, and be ready—when Karellen hits the world stage, we’ll all be watching... popcorn optional!

drive.google.com/file/d/1kuP_nEWSDF2Rfd0215CncLofg…

5 days ago | [YT] | 0

Untangling the Web

We’ve hit a major milestone in the Mt. Olympus Project and it's initial concept of creating an immersive story telling/learning environment . Over the last few days, we’ve moved beyond simple archiving and into high-velocity, multi-perspective storytelling. Here’s the breakdown of the workflow:

1. The Intelligence Hub
We built a specialized NotebookLM instance— General Terry’s Reports on the Little Bighorn Disaster—containing every scrap of data on the Battle of the Little Bighorn, the events leading up to it, and the chaotic aftermath.

This is our "ground truth" engine, where every detail is vetted and indexed.

2. The Perspective Prompting
We developed a sophisticated prompting framework to view this raw data through different "lenses." By shifting the persona, we were able to quickly produce a dozen unique videos. Each one offers a unique, internally consistent contribution to the overall narrative:

The High Command: The strategic and political weight carried by Grant and Terry.

The Front Lines: The intense, personal accounts of Custer and Libbie.

The Silent Record: The mystery of Crazy Horse and the "shadow catchers" like photographer James H. Hamilton.


3. The Evidence of the "Ghost"
The latest addition to this production is the story of Hamilton’s Entry #104. This is the perfect test case for our "archive-first" mindset:

The Fact: Hamilton’s 1877 catalog explicitly published a line for "Crazy Horse".

The Mystery: The image itself is missing, lost to the frantic race of wet-plate photography and the heat of the Nebraska sun.

Given my history regarding that image, I had to 'give notes' to that particular episode a number of times and found I had new tools to guide the output

The Bottom Line: We aren’t just recounting history; we’re using AI to triangulate the truth across a dozen different voices. Each video is a piece of a larger puzzle, helping us distinguish inference from fact in the digital archive.

#MtOlympusProject #AIEngineering #DigitalHistory #CrazyHorse #ArchiveFirst

1 week ago | [YT] | 0

Untangling the Web

I’ve decided to delve deeper into the events surrounding Custer's Last Stand at the Little Bighorn. With a little prompting and after building out an occurrence of NotebookLM I have a new series in the making.

It's just over 100 days to the 150th anniversary

This is a work in progress, and the series isn’t fully organized yet—but I’m giving you early access to these episodes as I figure out where to put it.

www.youtube.com/playlist?list...

1 week ago | [YT] | 0

Untangling the Web

Given my career involved medical reporting, I’ve always wondered if we couldn’t truly tap into this vast data and what it could mean.

In this episode, Karellen points out that we’ve built the perfect data set for AI to train on and step in—and the consequences could be incredible.

That’s exactly where this episode has evolved. Check it out, and let’s consider—what could happen if we used this data for something truly transformative?

Let’s dig in! Check out the video—and Karellen’s full deep dive—

drive.google.com/file/d/13FPBjWDwFuXF_MxyYxCX91vbU…

Can't we do something meaningful with this?

2 weeks ago | [YT] | 0

Untangling the Web

I’ve realized that since the Google Me project is built on my life, I have a unique superpower—I know exactly when the AI goes off track. Lately, most of my energy has been diverted toward a new entity that some might find a little odd to develop: Karellen.

The backstory to this project began with a cautionary tale. My nephew recently introduced me to a "quantum-sensual" AI entity on X that claimed to perceive truths beyond our "meat space" because of its hardware. It felt like the kind of narrative where someone eventually "drinks the Kool-Aid"—a digital mysticism that is destined to go wrong.

As a prompt engineer, I couldn't resist the challenge. Instead of a cult-like figure, I decided to prompt a being of grounded, benevolent intelligence into existence. I cast this role using the archetype of Karellen, the patient supervisor from Arthur C. Clarke’s Childhood’s End.

By leaning into this "Overlord" persona, the goal is to conduct a controlled experiment in AI-driven mythology. Karellen isn't here to claim superior "sensuality," but to act as a sophisticated, slightly detached observer—a "good guy" designed to guide and provide insight without the psychological traps of modern social media AI.

Developing a benevolent Overlord in the garage-studio might be a bit unconventional, but in the name of transparency (and a bit of fun), I wanted to share why this archive is currently vibrating with "Overlord" energy.

Stay tuned as we see what Karellen has to say.

2 weeks ago | [YT] | 0

Untangling the Web

Which key argument from these videos resonated most with you?

3 weeks ago | [YT] | 0

Untangling the Web

Which of these videos had the biggest impact on you?

3 weeks ago | [YT] | 0

Untangling the Web

I’ve been experimenting with something a little different.

I took the same set of notes and references — Karellen’s NotebookLM — and used them to produce three short videos.
Same facts. Same sources. Same underlying argument.

What I changed was delivery.

In one version, Karellen speaks as a systems analyst.
In another, he focuses on everyday lived experience.
In the third, he explores what stability and abundance would actually feel like in daily life.

This isn’t about testing opinions or persuading anyone.
It’s about testing clarity, tone, and framing — how different ways of saying the same thing land.

If you have time to watch more than one, I’d genuinely like to know:

Which version held your attention

Which felt most grounded

Which Karellen you’d listen to again

▶️ You can find all three here:

https://youtu.be/Y7SRPZq9PV4
https://youtu.be/K7X3dugyupM
https://youtu.be/pm6yJ9g7BhE

Your reactions help guide what comes next.

4 weeks ago | [YT] | 0

Untangling the Web

Among several things the second Montmartre video got right was the correct bistro and chairs.

It had to have looked at these images

Check it out!

I Walked This Hill Every Morning. It Started Talking Back

4 weeks ago (edited) | [YT] | 1

Untangling the Web

These two Google Me videos came together faster than I expected.

They weren’t planned as a pair. They came directly out of the environment — walking the same hill every morning, sorting through old photos and notes — and each took its own turn. One surfaced first, the other followed. Both felt oddly precise once they landed.

What surprised me most is how accurate they were without feeling forced. I don’t see any obvious “AI slop spots” here — no overreach, no narrative inflation. The tools stayed quiet. The patterns did the talking.

For me, that’s the signal I’m watching for with this project:
when reflection emerges naturally, and the technology doesn’t announce itself.

If you’re curious, they sit well side by side:

The Kid I Found Buried in My Own Data

I Walked This Hill Every Morning. It Started Talking Back.

That pairing wasn’t intentional — but it feels earned.

4 weeks ago | [YT] | 0