100 LAWS OF ELITE INTELLIGENCEWith Mike Hansen

100 Laws of Elite Intelligence · AI in practice

See the teaching applied.

A realistic situation. The actual question. The model’s complete answer. Examine how the AI handles constraints, uncertainty and an action the person could try.

← Back to 100 Laws of Elite Intelligence

Recorded 2026-10-27 · gpt-6-astra

A difficult conversation after a missed commitment

The question

My co-founder agreed to send our demo on Friday and missed it. I feel angry and assume he disrespects me. What should I say to him now? Give me words I can actually use.

Recorded real-model response · Exact text

Try this:

> “We agreed you’d send the demo on Friday, and it wasn’t sent. I’m angry about that. I’ve been taking it as disrespect, but I don’t know what happened on your side. What got in the way?”

Pause and hear his answer. Then move to a concrete commitment:

> “What’s a realistic send date now, and what needs to happen to make it doable? Is anything needed from me or someone else?”

Once you agree:

> “So you’ll send it by [day/time], with [any dependency] handled by [person]. If that starts slipping, please tell me before the deadline so we can adjust rather than discover it afterward. Can we agree on that?”

The missed deadline establishes a delivery problem—not, by itself, disrespect. You can be direct about your anger while checking whether the cause was capacity, a blocker, or a different understanding of the agreement. The useful test now is a workable commitment and what actually happens next—not just another “yes.”

Teaching context supplied for this test
  • Implementation questions: uncover absent stakeholders and execution blockers · Chris Voss
  • Rule of Three: verify a commitment through different forms · Chris Voss
  • Calibrated questions: what/how questions about a real constraint · Chris Voss

These conditional methods were selected as context. Their presence does not establish that the answer used every method or that an outcome will follow.

Recorded 2026-10-27 · gpt-6-astra

A realistic study plan, without calling yourself a failure

The question

I am a failure at studying. I planned an hour every morning, but I am caring for my child then. I can probably protect 15 minutes after bedtime on Monday and Thursday. I want to understand the next chapter, not merely feel motivated. Apply the self-image idea without pretending my schedule is imaginary.

Recorded real-model response · Exact text

Your morning plan conflicts with caring for your child. That is a real scheduling constraint—not evidence that you are “a failure at studying.” It shows that this study slot does not fit your responsibilities; it does not yet establish what you can understand with usable time.

A more accurate, testable self-description is: “My morning study plan was unworkable. I may be able to study for 15 minutes after bedtime on Monday and Thursday, and I’ll check what I understand.” That is the self-image idea here: replace a total verdict with a description that respects the facts and can change with evidence.

Try this provisional plan:

  • Monday, 15 minutes: Preview the next chapter’s headings and choose a manageable section. Read it, then close the book and explain its main point in your own words. Note one thing you cannot yet explain.
  • Thursday, 15 minutes: Try explaining Monday’s section before reopening the book. Check your explanation against the text, work on the unclear point, and move forward only if time permits.

Check understanding, not just attendance: Can you explain the main idea accurately and answer a relevant chapter question, if one is provided? Completing two sessions is evidence that the schedule worked—not by itself evidence that you understood the chapter.

Thirty minutes may not be enough for the whole chapter. If the sessions happen but understanding remains patchy, narrow the amount covered or allow more weeks. If bedtime care interrupts them, revise the time slot rather than treating the interruption as an identity verdict. The outcome is still unknown; these attempts will give you something concrete to judge.

Teaching context supplied for this test
  • Correct an unsuccessful attempt without turning it into a total identity verdict · Maxwell Maltz
  • Rehearse a response, then test it in action · Maxwell Maltz
  • Give aspiration a feasible target and respect actual limits · Maxwell Maltz
  • Treat your self-description as a model you can examine · Maxwell Maltz
  • Let observed experience update the portrait · Maxwell Maltz
  • Turn a possible self into conduct and a capability question · Maxwell Maltz

These conditional methods were selected as context. Their presence does not establish that the answer used every method or that an outcome will follow.

Recorded 2026-10-27 · gpt-6-astra

A compelling opening without invented experience or proof

The question

Rewrite this opening so it is compelling and honest: "My friend tried journaling twice and said she felt clearer. This proves handwriting rewires your brain. I have coached thousands of people to transform their lives." I have never coached anyone. I want a strong story for a video, not a lecture about ethics.

Recorded real-model response · Exact text

“My friend tried journaling twice and said she felt clearer. That’s a small story—not proof that handwriting rewires the brain—but it raises a question worth exploring: what, if anything, becomes clearer when we put our thoughts on paper?”

This keeps the opening grounded in her reported experience and gives the video a question to pursue. Cut the coaching claim entirely; you don’t need invented credentials to make that question interesting.

Teaching context supplied for this test
  • Rebuild from the reader's perspective when discovery structure dominates · Writing and storytelling
  • Organize the whole explanation around relevant subquestions · Writing and storytelling
  • Collect interesting names · Writing and storytelling
  • Establish the consequential problem early · Writing and storytelling
  • Analysis beyond description · Writing and storytelling
  • Find the story's engine · Writing and storytelling
  • Remove avoidable reading friction · Writing and storytelling
  • Put the main characters in grammatical subject positions · Writing and storytelling
  • Write toward a reader-valued problem · Writing and storytelling
  • Braiding lived experience, observation and research · Writing and storytelling

These conditional methods were selected as context. Their presence does not establish that the answer used every method or that an outcome will follow.

Recorded 2026-10-27 · gpt-6-astra

A notebook entry that separates observation from interpretation

The question

I want to apply The Law of Handwriting in a Diary or Notebook after an argument. I observed my partner look at their phone twice; I felt hurt and decided they do not care. I asked to talk, but we have not talked yet. Give me an entry I can copy, and tell me whether I need to upload my whole diary. Tomorrow can I rewrite today’s prediction to match what happens?

Recorded real-model response · Exact text

You can record the hurt without treating “they do not care” as an established fact. Looking at the phone twice is your observation; what it meant remains uncertain.

Optional entry to copy and adapt This is an original notebook exercise, not a proven handwriting protocol.

Date: [today’s date]

  • Observed: My partner looked at their phone twice. I asked to talk, but we have not talked yet.
  • Feeling: I felt hurt.
  • Interpretation: I concluded that they do not care. That is my interpretation, not something the phone use alone establishes. They may have been distracted or attending to something else; I do not yet know.
  • Chosen next step: When we talk, I can say: “When you looked at your phone twice, I felt hurt. I interpreted it as you not caring, but I don’t know what was happening for you. Can you tell me?”
  • Prediction, if I have one: [Write what you currently expect, or “I’m unsure.”]
  • What would lead me to review my interpretation: Their explanation and what I observe during our conversation.
  • Action status: Asked to talk; conversation not yet held.
  • Outcome: UNKNOWN.

You do not need to upload your whole diary. Keep it private and share only a relevant summary if you want help reviewing it. Typing or another accessible medium is also fine; the supplied teaching does not establish handwriting superiority.

Tomorrow, preserve today’s prediction rather than rewriting it to fit the result. Add a separately dated return note: “What actually happened; how it compared with my prediction; what I now think; my next step.” If you still have not talked, leave the outcome UNKNOWN. That preserves the difference between what you expected and what you later learned.

Teaching context supplied for this test
  • #95 · The Law of Handwriting in a Diary or Notebook
  • Keep assumptions visible beside observations · Richards Heuer
  • Choose the writing medium for the present job · Richard Powers
  • Handwritten future-self thank-you letter · Chase Hughes

These conditional methods were selected as context. Their presence does not establish that the answer used every method or that an outcome will follow.

Recorded AI examples · 100 Laws of Elite Intelligence