How To Use AI Better Than 99% Of People (3 mental models)

Key Concepts

• Big Boss Energy: A mindset shift from using AI as a mere assistant (task-based) to a Chief of Staff (strategic partner).
• Higher-Order Thinking: Moving up Bloom’s Taxonomy by using AI for synthesis, evaluation, and creation rather than just rote task completion.
• Loop Thinking: A methodology where processes are iterative, incorporating feedback to improve outcomes rather than following a linear "completion logic."
• Behavioral Change: The ultimate goal of learning; if no behavior changes, no actual learning has occurred.
• AI Slob: The phenomenon of generating low-quality, high-volume content due to a lack of strategic oversight.

1. The Shift: From Assistant to Chief of Staff

The speaker argues that seniority is no longer the primary metric for success; the new differentiator is how effectively one leverages AI.

• The Assistant Model: Treats AI as a tool to manage calendars and basic tasks. It provides leverage but keeps the user in a reactive, "last in rank" position.
• The Chief of Staff Model: Positions AI as a thought partner. This agent manages other agents or workflows, helps map out priorities, and turns high-level visions into executable plans.
• Strategic Advantage: By adopting "Big Boss" energy, you are not just delegating tasks; you are managing a 24/7 team of AI agents, allowing you to perform at an executive level regardless of your formal job title.

2. The Three Pillars of "Big Boss" Workflows

To operate effectively with AI, the speaker proposes a framework where time is divided into three core processes:

A. Learning (Behavioral Change)

• The Process: Learning is not just consuming information; it is a cycle of input, behavioral change, and feedback.
• Methodology:
  1. Consume input.
  2. Map it to a specific situation.
  3. Change one behavior.
  4. Analyze the feedback to refine the next input.
• Key Insight: If your behavior remains the same after "learning," you have not actually learned anything.

B. Creating (Problem Ownership)

• The Process: Creativity is a proxy for intelligence. Instead of being a "cog" executing someone else's solution, you must identify the problem yourself.
• Example: The speaker describes separating "Idea Vaults" from "Practical To-Do Vaults" in Obsidian. By identifying that these require different cognitive modes (divergent vs. convergent), the user creates a custom system that solves their specific workflow friction.

C. Systematizing (Scaling via SOPs)

• The Process: Making knowledge explicit so that AI agents or human teammates can execute tasks at your standard.
• Methodology:
  • Breakdown: Deconstruct vague tasks into granular steps.
  • Emphasis & Leverage Points: Identify the "80/20" of the process. For example, in document creation, emphasize "quality checks" (numbers, dates, fonts) and "data integrity" (garbage in, garbage out).
  • Skill Building: Use AI to codify these steps into "skills" that the agent can reference, preventing the need to start from scratch every time.

3. Loop Thinking vs. Completion Logic

• Completion Logic: The traditional, low-level approach of "I have a task, I finish it, I move on." This is how interns or junior staff often operate.
• Loop Thinking: An iterative approach where the output is constantly refined.
• Real-World Application: The speaker cites Japan Airlines as a model for "passing the baton." Each team (ground staff, flight attendants, etc.) anticipates the needs of the next team in the chain, creating a continuous loop of quality improvement.
• AI Integration: Modern AI agents (like those using Claude Code or Auto Research) function best when they operate in loops—planning, executing, checking, and refining—rather than just providing a single, static response.

4. Quality and the "Big Boss" Perspective

When AI makes output cheap, quantity is guaranteed, but quality must be curated.

• Three Drivers of Quality:
  1. Better Exploration: Using AI to conduct research at a scale impossible for humans.
  2. Better Loops: Using feedback mechanisms (like Andre Karpathy’s "Auto Research" concept) to iterate toward better results.
  3. Better Criteria: Applying your own taste, judgment, and point of view to the AI’s output.

Synthesis and Conclusion

The core takeaway is that AI technology is a mirror: if you use it for low-level tasks, you remain a low-level operator. If you use it to exercise your own point of view, you become a strategic leader. By shifting from "getting things done" to "owning the problem and the process," you reclaim your cognitive capacity. The goal is to move away from the "AI slob" mentality of mindless generation and toward a high-order, loop-based system where your unique judgment dictates the quality of the output.