Modus Promethean's
Unleashing AI

Success in the next decade belongs to the people who learn to leverage their own skills with AI. Everyone else gets replaced by it.

If you're measuring what AI does for you in percentages — 30% faster, twice as productive — you're still thinking in the old frame. Think multiples. Not of speed, of what you can now attempt.

Something worth asking first

What you leave with

• You will have personally directed a live AI agent to build something — not watched a recording of someone else doing it.
• You will have seen an agent design its own working memory, unprompted, and understand why that's the actual unlock, not a party trick.
• You will have pulled real data yourself and built your own analysis from it — the difference between asking a chatbot for an answer and checking one yourself.
• You will know the one habit that separates people who get real leverage from AI from people who get frustrated by it — and it isn't a better prompt.

Two tracks, so you start in the right place

Both run live over Zoom, about four hours. Pick your track when you register — a few short questions route you automatically.

Who's teaching it

Thirty-plus years in software engineering, most of it on the infrastructure and support side of the business, not just the code. Six years on Wall Street as a VP running infrastructure and support — including managing the disaster recovery effort after September 11, in the trenches himself as much as running it from above. Then: a bookkeeping company he ran himself, software built for government contractors and a handful of niche services, a PC configuration center, a call center. Different rooms, different industries. Same discipline every time — the system stays up, the queue clears, the client gets an answer.

The last year has been spent almost entirely inside AI systems — not reading about them, running them: finding the fastest reliable way to produce consistent code, images, documents, and working web pages on demand, and testing where that reliability actually breaks. Before that, testing and training AI models directly for a data-labeling firm — the unglamorous work of telling a model what "correct" means, thousands of times, until you understand its failure modes from the inside instead of the outside.

He doesn't perform expertise. He gets to the point with minimal translation, and expects the same back from the tools he uses. This seminar is that standard, taught directly.

In his own words

This isn't a curriculum reverse-engineered from a pitch — it's drawn from what he's already published, in public, dated. He also writes for Covenant, a longer-running project on AI governance and protocol, if you want the fuller picture beyond this seminar.

• If You're Measuring AI in Percent, You're Not There Yet Why "30% faster" is the wrong way to think about what's possible now.
• When the Levee Breaks What token caps tell you about pressure that's been building longer than anyone's been counting.
• Be Your Own Corporation The thing a job actually gave you wasn't the paycheck. AI lets you build that yourself.
• You Sound Like an AI The difference between people who get sharper using AI and people who don't isn't the tool.
• It Isn't the Model. It's the Scaffolding. Why the same instruction gets sharper results in one tool than another — and it isn't the model doing it.
• Your Chatbot Isn't Thinking About You Right Now What's actually happening between your messages. Less than you think, and it changes the whole question.
• The Opus 5 Debate Is One Layer Too High A different lens on where the real argument about AI capability actually lives.