The Notebook Technique: Structured Thinking with Artificial Intelligence

The notebook technique fundamentally changes the way we work with large language models (LLM). Unlike traditional prompting, where we simply state our request and hope for the best, the notebook technique creates a structured thought process that both you and the LLM can follow.

Consider this common scenario: When asking an LLM to analyze customer buying patterns, a typical challenge might simply ask for the analysis directly. Result? Often a rushed or ill-advised response that misses key considerations. This is where the notebook technique proves to be transformative.

Instead of diving directly into the solution, the notebook technique introduces a critical planning phase. By using tags, you create a space where LLM can:

  • Map his approach before execution
  • Break down complex problems into manageable parts
  • Identify potential pitfalls early
  • Expect a course correction before investing time in the wrong direction

Real power lies in transparency. By seeing how the LLM plans to deal with your application, you can immediately catch misunderstandings and direct it to the intended outcome.

Create your first Notebook

The basic structure of a notebook challenge consists of three key elements:

  1. Initial requirement: Present your task clearly, but don’t stop there. Include context about your goals and any specific requirements.
  2. Scratchpad Guidelines:Direct the LLM to plan its approach using brands. Be specific about what you want to see in the planning phase.
  3. Review loop: Before proceeding with a full answer, please examine the proposed approach and refine it as necessary.

Let’s go through a practical example. Instead of just asking for a market analysis, you can structure it like this:

A basic notebook response will look something like this:

This planning phase reveals the intended focus areas of the LLM. If you notice that it’s missing key elements—perhaps an innovation in battery technology or a partnership with charging infrastructure—you can be redirected before receiving a complete but off-target analysis.

Common pitfalls to avoid

  1. Wave Planning Guidelines: Don’t just ask the LLM to “plan your approach”. Request specific elements you want to see in the notebook, such as methodology, data points, or analytical frameworks.
  2. Skip the check step: A notebook isn’t just for show—it’s your chance to course correct. Take the time to evaluate whether the proposed approach meets your needs.
  3. Insufficient context: While a notebook helps organize thinking, it cannot compensate for unclear initial requirements. Be specific about your needs, audience, and intended use of the information.

Beyond the basics

When tackling LLM projects, creating the right opening challenge becomes essential.

Here are some quick ideas for layered challenges. Be sure to include as much context as possible in the form of attached documents or text:

Product launch strategy challenge:

  • “I need to develop a comprehensive product launch strategy. Using tags compartmentalize your approach to market position analysis, competitive landscape, and go-to-market planning. Consider potential information gaps and dependencies between these elements before making any recommendations.”

Budget Planning Challenge:

  • “Help me create the departmental budget for next year. Use markers before entering numbers to outline your methodology for considering historical data, growth projections and resource allocation. Mark all the prerequisites you will need in this process.’

Call for Project Risk Assessment:

  • “We are launching a new enterprise software system. Before listing potential hazards, use labels map out how you will approach risk identification across technical, operational and business dimensions. Include your framework for prioritizing these risks.”

Each of these challenges:

  • It clearly sets the end goal
  • It requires specific planning elements
  • Identifies key areas requiring attention
  • It encourages systematic thinking before generating solutions

Note: These prompts are fairly simple examples. You can get very detailed and complex depending on the use case.

The key is to be explicit that you want to see the planning process before you get the solution. This prevents hasty reactions and ensures that all relevant factors are carefully considered.

As already mentioned, these challenges can (and should) be further adapted:

  • Adding specific industry context
  • Including known limitations
  • Specification of desired outputs
  • Please indicate the specific methodologies you wish to consider

Remember: The more complex the task, the more important it is to see the LLM’s planned approach before it generates a solution.

Develop your rapid engineering skills with

Learn challenges

The notebook technique helps us transform the interaction between people and LLM. By implementing this structured approach, organizations and professionals can go beyond simple patterns of rapid response and achieve more sophisticated and reliable results. This methodology is especially proven in environments where accuracy, consistency and transparent reasoning are paramount to decision-making processes.

Leave a Comment