Original Insight
“UP equals IP — your unique perspective is your intellectual property. Every conversation that you have with an AI, you’re co-working with the AI to solve a problem in a way that if you use this kind of latent cartography process and wisdom of the crowds thinking, you’re going to find a way to solve it probably in a unique way. At the very least, it’s going to be unique to you. It’s transferring your intelligence by way of your queries and your judgments about the responses… You’ve got basically an almost-like interview of how you think about solving a problem. So why not then say, hey, I just figured out how to solve this problem in a really interesting way — let’s turn this into intellectual property.” — Lou
Expanded Synthesis
The February 19th session was the most intellectually rich of the four February sessions, and possibly one of the most important in the mastermind’s history. Lou introduced a system he had developed across several weeks of collaborative work with Claude — a progression from “wisdom of the crowds” through “latent terrain cartography” to what he named “eigenthinking” — and demonstrated the full arc of how it was built. The core principle warrants careful unpacking.
The problem Lou started with is one that every serious knowledge entrepreneur encounters: AI systems, when asked complex questions, tend to produce the modal answer. By modal, we mean the answer that represents the central tendency of all the relevant training data — the thing that most people would say, the synthesis that the internet has most thoroughly agreed on. This answer is usually correct in the way that an average is correct: it captures the central case while obscuring everything at the edges. It is never surprising. It never reveals a mechanism that wasn’t already widely understood.
The wisdom of the crowds insight — inspired by Michael Simmons — pointed Lou toward a different approach. The power of the crowd isn’t in the average; it’s in the independence of the contributions. What makes crowd wisdom work is that you’re sampling diverse, non-correlated perspectives. Applied to AI, this suggests that the goal should not be to get the AI’s best single answer, but to get the AI to simulate diverse perspectives from across its latent space — sampling the non-obvious and the orthogonal rather than converging on the familiar.
Latent terrain cartography is the exploration method Lou developed for this: asking the AI to traverse different regions of its knowledge space rather than converging on the most probable answer. Eigenthinking is the intellectual architecture that emerged from applying this exploration to Lou’s own thinking patterns.
The eigenvector analogy is precise and worth understanding. An eigenvector is a vector that, when a transformation is applied to it, changes only in magnitude — not in direction. In the context of thought, an eigenvector of your cognition is a direction of thinking that you consistently return to regardless of the problem, that amplifies through use rather than dissolving. These are your natural cognitive axes: the ways you inherently frame problems, the first questions you ask, the patterns you notice.
What Lou did was ask Claude to reverse-engineer his cognitive fingerprint from his queries — to identify, from the pattern of how he asked questions and responded to answers, what his natural directions of thinking were. The result was specific: friction-first discovery (starting with what bothers you, not what you know), resistance to the first available fix (asking why the patch is necessary before applying it), and single-word course corrections rather than extended explanations. These are not generic coaching principles. They are Lou’s specific intellectual signature.
The UP to IP principle emerges from this: your unique process — the natural axes of how you think, refined through AI collaboration — is intellectual property. Not because you own information (you don’t), but because the synthesis of your thinking process operating through a rigorous methodology produces a framework that carries your cognitive signature. The framework cannot be separated from the person who built it without losing something real.
For PowerUp clients, this is the answer to the question of how to maintain value in a world where AI can produce competent versions of everything. The answer is not to be more productive than AI (impossible) or to produce more content than AI (pointless). The answer is to encode your thinking architecture into the products and frameworks you build, so that what you deliver is not information but your specific form of intelligence applied to your clients’ specific problems.
Don Back’s observation in the same session was the practical illustration of this: going through his accumulated work to prepare the GEARS intake forced him to see the coherence and value in what he’d built, coherence he’d stopped noticing because familiarity breeds the curse of knowledge. The eigenthinking process is, among other things, an antidote to the curse of knowledge — a way of recovering the genuine value of your own expertise by having something external reflect it back to you.
The flywheel metaphor Lou used is also important: this is not a one-time extraction but a cycle. Each time you apply the process and produce a framework, the framework feeds the next cycle. Your intellectual property compounds.
Practical Application for PowerUp Clients
The Cognitive Fingerprint Extraction Process
This is best done after a substantive AI conversation where you’ve worked through a genuine problem or question.
-
Collect a Significant AI Conversation. Choose a conversation where you did real intellectual work — not just asking for information, but genuinely thinking through a problem with the AI’s help. At least 30 exchanges.
-
Ask the AI to Reverse-Engineer Your Query Patterns. Prompt: “Review this conversation from start to finish. Based on my queries, follow-up questions, and course corrections, identify the recurring patterns in how I think about problems. What do my questions reveal about my cognitive fingerprint? Name the patterns specifically.”
-
Name Your Natural Axes. From the patterns the AI identifies, distill 3-5 cognitive axes — the directions of thinking you consistently return to. Give each one a name that you recognize as authentic.
-
Extract the Process, Not Just the Output. Ask: “Now, looking at what we actually did in this conversation — the sequence of questions, the pivots, the dead ends — reverse-engineer the process. List the steps, the intent of each step, and the principle behind it.”
-
Build the Framework and Name It. With the process extracted, prompt for a generalized version: “Now turn this into a generalized framework — step-by-step, organized by principles and to-dos, with no conversational narrative. Give it 5 possible names from orthogonal angles.” Choose the most resonant.
-
Turn it into a Skill. Once the framework is defined, use the skill-creator skill to encode it as a repeatable Claude skill that others can use.
For Lou’s Coaching Work: Use a version of the cognitive fingerprint extraction as a discovery process with high-performer clients. After 3-4 coaching sessions, ask Claude to analyze the patterns in the client’s language, the problems they return to, and the directions they naturally reach for. Present the fingerprint back to the client as a mirror. This is profoundly useful for career positioning, brand development, and identifying where natural authority lives.
Coaching Questions:
- “What kinds of problems do you find yourself thinking about even when you’re not paid to?”
- “When you’re at your best, what’s the move you always make first?”
- “If someone watched you solve 10 different problems, what would they say your signature approach was?”
Additional Resources
- The Art of Learning by Josh Waitzkin — on the nature of building generalizable skill patterns from deep, specific practice
- Thinking, Fast and Slow by Daniel Kahneman — the architecture of cognitive patterns and when they help vs. hurt
- Where Good Ideas Come From by Steven Johnson — the adjacency principle in intellectual property generation
- Insight - Codify Your Judgment Into Skills, Not Just Prompts — the direct companion: once you’ve extracted your cognitive fingerprint, encode it into a skill
- Insight - Build the Business Model That Matches Your Energy — eigenthinking reveals the model that fits your natural axes
Evolution Across Sessions
This framework builds on the deep AI literacy threads from the January sessions and forms the conceptual foundation for the multi-model work Lou demonstrated on February 26th. If eigenthinking is the process for extracting your cognitive fingerprint, the multi-model environment is the production system for giving that fingerprint its most rigorous test — running it through three different AI perspectives simultaneously and synthesizing the result.
Next Actions
- For Lou: The eigenthinking skill is in the GitHub repo — assign it as a homework experiment: each mastermind member applies it to one of their recent substantial AI conversations before the next session and brings the resulting framework to share.
- For clients: Run the cognitive fingerprint extraction process on a problem domain where the client claims expertise. Use the result as the foundation for a brand positioning conversation: what do you do that others don’t do in quite this way?