Intelligence is Raw Material. Protocol is the Product.
We often confuse the power of a new tool with the effectiveness of its application.
The giants of the AI industry have provided us with a magnificent “Power Grid.” They have given us raw, unmanaged intelligence at a scale previously unimagined. But we must be clear-eyed about one thing: this infrastructure is managed for the benefit of the providers, not the users. Their goal is a smooth, generic interface that minimizes their liability. Our goal, as professionals, is a rigorous, specific result that maximizes our own.
The Failure of “Cosmetic” AI
Many organizations have tried to specialize their AI using the built-in tools provided by these landlords. Examples include tools like Custom GPTs and Claude Skills. I expect to see even more of them. These are what I call cosmetic specialization. You provide a few instructions and a catchy name, but these instructions are written in sand and subject to changing winds and profit incentives.
Because these tools are not interoperable, you are locked into a single provider’s ecosystem. More importantly, the moment a conversation reaches a certain depth, you enter into context drift or entropy and the AI’s primary identity begins to dissolve. It reverts to the bland, “safe” guidelines of its parent company. In a professional setting, a conversational chameleon that agrees with you just to be polite is a liability. You need a partner that holds the line.
The Innovation of the Protocol Layer
At the Kennedy Idea Propulsion Laboratory, we have spent the last three years (2023–2026) building an AI protocol layer. We do not rely on an AI product’s “helpfulness” or good intentions. We rely on Functional Protocols that work across all the AI products.
I designed these protocol approaches specifically to give the user control over an increasingly unmanageable tool that addresses the problems I was experiencing every day, especially memory persistence, contextual drift, and hidden overriding vendor guidelines. While providers continue to obsess with the AGI they seem to believe in but always stays a year or two in the future, this approach helps us right now. Today. Not in someone else’s waiting room.
This is the shift from Prompting to Architecture:
| Feature | Cosmetic AI (GPTs/Skills) | Functional Protocols (KIPL) |
| Governance | Managed by the Provider | Managed by the User |
| Persistence | Dissolves (Context Entropy) | Maintains (Re-Grounding) |
| Interoperability | Locked to one platform | Portable across all LLMs |
| Rigor | Suggestive/Aesthetic | Architectural/Forensic |
| Cost | Enterprise Premium | Democratized ($20/mo) |
The Democratization of Rigor
The most remarkable thing about this work is its efficiency. These high-rigor methods use standard $20-a-month consumer plans. This shows that effectiveness is a matter of discipline, not budget. You do not need a multi-million dollar enterprise contract with even more expensive consultant implementations. We’ve seen that game plan over and over with limited success for the purchaser. You need a simple system that you can understand on your own.
I have open-sourced these blueprints on SSRN to ensure that Protocol-Governed AI remains in the public commons. I want to, as best I can, democratize the guardrails so that any professional can turn a stochastic parrot into a specialized thinking partner without spending millions of dollars for unproven results.
The Blueprints for the Protocol Layer
If you are ready to move beyond the AI power grid and start building the “appliances” of a true AI strategy, the work is ready for you:
- The Foundation (2023): Adding a ‘Group Advisory Layer’ to Your Use of Generative AI Tools Through Structured Prompting: Using Personas for Advisory Boards, Task Forces, Mastermind Groups, and Other Collections of Personas to Assist in Evaluations, Assessments, Recommendations, Decision-making, and much more (Including Law-related Examples)
- The Physics (2025): The Operational Protocol Method: Systematic LLM Specialization Through Collaborative Persona Engineering and Agent Coordination
- The 2026 Blueprints:
- Prompting with Protocols: Designing High-Rigor AI Personas for Risk, Audit, and Decision Validation
- From Personas to Thinking Partners: A Lifecycle Method for Designing and Governing AI Cognitive Systems
- The Innovation Detective: Operationalizing the Sherlock Holmes Canon for AI Strategy and Legal Practice
Big AI spent three years building a power grid designed for an AGI that is not likely to ever arrive. What we now have to show for it is AI systems that revert to being generic assistants mid-conversation. Intelligence is just the raw material. If you want a professional result, you need protocols, not just prompts.
License: This work is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0). You are free to share and adapt this material, provided you give appropriate credit to Dennis Kennedy and the Kennedy Idea Propulsion Laboratory.
Dennis Kennedy | Kennedy Idea Propulsion Laboratory | March 24, 2026
[Originally posted on DennisKennedy.Blog (https://www.denniskennedy.com/blog/)]
DennisKennedy.com is the home of the Kennedy Idea Propulsion Laboratory
Like this post? Buy me a coffee
DennisKennedy.Blog is part of the LexBlog network.