PowerUp Coaching — Living Knowledge Base

Let AI Watch You Work — Then Ask It What It Noticed

5 min read

Topic

Using AI as an NLP-style observer of your own working process — letting you do real work naturally, then asking Claude what it noticed about how you think — as a method for surfacing tacit expertise you’ve never been able to articulate or codify.

Target Reader

Experienced coaches, consultants, and knowledge entrepreneurs who feel their expertise is hard to transfer, package, or scale because so much of it is intuitive — they “just know” but can’t explain why. They’re already using Claude but haven’t thought of it as a mirror for their own cognition.

The Fear / Frustration / Want / Aspiration

Frustration: “I know I’m good at this, but I can’t explain exactly why — which means I can’t teach it, can’t delegate it, and can’t build AI that replicates it.”

Aspiration: A way to get the unconscious parts of their expertise out of their head and into a form that’s teachable, scalable, and embeddable in AI workflows — without having to sit down and try to introspect on demand.

Before State

The reader tries to codify their expertise by sitting down and writing it out — but the blank page is hard. They get the obvious stuff, but the real edge — the thing they do differently from everyone else in their field — stays locked inside because they can’t see it from the inside.

After State

The reader has a process for letting their tacit expertise surface naturally. They work on real problems, Claude observes, and then they review what Claude noticed. The surprising observations become the seeds of their most distinctive IP. Over months, a profile builds that reflects not what they think they know, but what they actually do.

Narrative Arc

You’ve spent ten years getting good at something. You know things you can’t explain. When a colleague asks “how did you know that?” your honest answer is “I just did.” NLP practitioners have always known how to mine this — they watch experts work and notice what the expert can’t notice about themselves. The twist is that you now have a language model that can do this too. You do the work. Claude watches. Then you ask it: what did you notice? The answer will surprise you.

Core Argument

Your best expertise is invisible to you — but not to a model that watched you use it. The question “what did you notice about how I approached this?” unlocks tacit knowledge that no amount of introspection can.

Key Evidence / Examples

  • Direct quote from source insight: “It’s like NLP — you watch the expert do their thing, and then you notice some things they’re not even conscious of, like: why did you grip the pistol like that? Why did you stop your breathing right there? So that’s kind of what I’m trying to get Opus to do on me.” — Lou
  • The NLP parallel: Bandler and Grinder’s foundational work modeled Erickson, Perls, and Satir by watching them work — not by asking them to explain. The observation precedes the articulation.
  • Supporting insight: Insight - EigenThinking — Turn Your Cognitive Fingerprint Into Intellectual Property — the broader framework this technique serves
  • Contrast with deliberate codification: the top-down “write your frameworks” approach misses the tacit layer; this bottom-up approach gets it

Proposed Structure (5–7 beats)

  1. The invisible expertise problem: What you know in your hands vs. what you can say with your mouth — and why the gap matters for coaches and knowledge entrepreneurs.
  2. What NLP discovered about expert modeling: You can’t get tacit knowledge by asking for it. You get it by watching and noticing.
  3. AI as observer: How to use Claude not as a collaborator but as a witness — and what changes when you ask it what it noticed rather than what you said.
  4. The protocol in practice: Work naturally → ask “what did you notice about how I approached this?” → confirm, correct, or probe each observation.
  5. The cumulative profile: How repeated observer sessions build a living document of your cognitive fingerprint — one that deepens over time.
  6. From observation to operationalization: How to take what Claude noticed and embed it in prompts, skills, and AI workflows — so your tacit expertise becomes executable.
  7. The deeper implication: You’ve always had this expertise. What’s new is that it no longer has to stay invisible.

Editorial Notes

  • Tone: Intellectually curious, slightly philosophical — this is one of those ideas that makes you stop and think “I could actually do this today.” Avoid making it sound mystical or overclaiming on what AI can observe.
  • The NLP connection is strong but niche — introduce it gently, don’t assume the reader knows it, and don’t over-explain it.
  • The actionability score is slightly lower (3.5) because the protocol requires genuine unguarded work, which is harder to manufacture on demand. The article should be honest about this — don’t make it sound like a 5-minute exercise.
  • Competing territory: “AI for self-reflection” content exists but is typically journaling-focused. This article should differentiate by being about work, not about feelings.

Next Step

  • Approved for drafting
  • Needs revision
  • Deprioritised

Graph View

Backlinks