What do you do with the information that you glean from books? If you’re like most people, you read some great books, highlight them, take notes, and then put them on the shelf to gather dust.
Mark Cuban once said that a $20 book is one of the best investments that he can make because of its expert wisdom. He reads three hours a day just to learn what others are thinking. So, how do you put that information to use? Do you take the lessons and update your processes? In the age of AI, you can incorporate lessons learned and immediately gain benefits. You can apply lessons learned to your discipline, business, or even hobbies.
Moving Beyond Traditional Prompting
You can add lessons learned through traditional prompting. But this does not scale well, and you’re missing out on business context. There is a different approach that scales. That approach is using AI priming documents. AI priming documents are separate documents that can be integrated into your GPTS or prompts to add context or additional instructions.
We want the output to reflect who we are and what our business does, while also incorporating new information we’ve gleaned. This involves context engineering. This involves including context about ourselves, our business, our services, our branding, and our core tasks.
AI Priming Documents
This post focuses on improving your AI processes based on the lessons you learned. However, AI priming documents are essential to the conversation. In brief, they are standalone documents that you include in a GPT or prompt as appropriate to the task or purpose. Some of the documents I recommend include:
- Professional profile
- Business/institution/department profile
- Service profile
- Branding guidelines
- Report/formatting examples
You can learn more in my free book, AI for Learning Leaders.
Improving Your Processes
Another set of AI priming documents is your business processes. You can build a single document or create separate documents if you want to separate functions, e.g., marketing, publishing, content creation, etc.
Creating separate documents will make updates easier. This is where squeezing the essence from what you read comes into play. Actively engage with your reading to find the processes, tips, and tricks that will improve your business.
Focus on the “how” and the “why” of the book’s logic. Pull out that expert-level context that you can use to improve your AI logic. You have to distill the information you extract from a book. Make it leaner than just uploading the whole book.
Your generative AI tools can help you experiment and refine it before you update your process. Naturally, you will need to thoroughly test the changes before incorporating them into your system.
Your AI system is a collection of documents. I personally use Google Docs. Only use documents that enhance your AI GPTs and prompts.
The AI Layered Cake: Building Your Context Stack
Once you’ve distilled the information and added it to your AI priming document, you’re going to build your context stack, an AI layered cake.
Layer 1: The Base Context
The first layer is about you and your business: your personal profile, professional profile, business history, and your purpose. This includes the essence of your business, your vision, and your mission statement. This layer tells the AI who you are, what your goals are, and how you do business.
Layer 2: Operating Frameworks
The next layer consists of your operating tasks and frameworks. This covers how you handle finances, marketing, and sales. A great place to start is documenting your processes in themed files that you can feed into Gemini when building your “Gems” or ChatGPT to build GPTs.
Layer 3: Project-Specific Functions
The third layer is for project-specific tasks. This is where you create a unique GPT or tool to carry out a specific function, such as evaluating book chapters, analyzing proposals, or creating social media content. These tools use the previous layers (the priming documents) to inform how the AI responds, adding the specific “flavor” or core messages you want to bring forward.
Operationalizing Information
Having this modular approach allows you to pull in different elements as needed for expert-level results. When you read something new, think about where it fits:
- New Methods: It could be a new way to do something.
- Improvements: It could be an improvement on an existing process.
- Domain Context: It could be information relevant to a specific domain or institution you work with.
By updating your priming documents and GPTs, your AI learns and improves alongside you.
A Living Process
This is a living process. Conduct quarterly reviews—perhaps using OKRs—to ensure your processes accurately reflect your current learning and business model.
The beauty of these AI priming documents is that they are portable. Major models like ChatGPT, Gemini, and Claude can all use them, allowing you to leverage the best AI package for the task while maintaining your core contextual information.
Your Next Step
As a reader, what are you doing with what you learn? Are you putting the book back on the shelf, or are you operationalizing that information?
With the next book you read, I want you to:
- Squeeze the essence out of it.
- Update your AI priming documents (or create one if you haven’t yet).
- Capture procedural tips that help you do what you do better.
I would love to hear how you make out on it.
