In my previous article, “Crafting an AI sovereign solution for technical writers“ (tcworld magazine, January 2025), I introduced the concept of a sovereign AI solution, an AI tool that has been developed and customized to meet an organization’s unique documentation needs. While highlighting the potential for improved accuracy, efficiency, and innovation, AI design can be enhanced further by adopting and integrating the cognitive principles, which were introduced by Professor Kirk St.Amant during his presentations at MEGAComm 2023 and 2024.
St.Amant’s presentations, The Cognition of Content Creation for Usable Designs [1], and Cognition, Communication, and Commerce: A Communication-Based Approach to New Product Development and Design [2], introduced a framework for designing sovereign AI solutions that align with users’ reflexive, behavior-driven responses. By weaving these approaches into the development of an AI solution, organizations can create documentation tools that align with users’ intuitive interactions and cognitive patterns.
Cognitive reflexes
The concept of “behavior as reflex” explains how users interact with content based on learned patterns. For example, seeing a familiar stop sign prompts users to hit the brake intuitively and automatically. This reflexive behavior is a motivator for designing intuitive technical documentation that seamlessly guides users through tasks in a natural manner, striving to eliminate the cognitive strain.
Application in AI
In a sovereign AI solution, incorporating reflex-based design can create intuitive prompts and workflows. By using St.Amant’s principle of creating familiar “scripts” (predictable behavior patterns based on environmental cues), the AI can offer content elements that reflect user expectations such as logical flow, relevant diagrams, and consistent terminology.
The 3Cs of usable design
St.Amant’s 3Cs model for usability emphasizes three main components: context, cognition, and communication. According to this model, usability is determined not just by the content, but also by how well it fits the user's situational context, mental state, and preferred modes of understanding.
Application in AI
Contextual relevance
A sovereign AI solution’s tools need to be customized to an organization’s specific documentation environment. St.Amant’s insights on usability as a combination of user, context, and state (emotional or cognitive) provide a deeper layer to this customization. For instance, an AI system can suggest contextual cues or relevant information based on user scenarios, whether they’re troubleshooting, exploring a product feature, or performing routine maintenance.
Cognitive fit
The sovereign AI’s recommendations should align with cognitive ease, helping users to navigate complex information without overwhelming them. St.Amant’s principle of “replicating the design of space” (designing content environments that align with familiar user scripts) means structuring AI recommendations to follow user-intuitive steps, using familiar language and guiding transitions that reduce the mental load.
Communication patterns
Effective communication involves presenting content in ways that align with how users consume information. A cognitive-focused sovereign AI solution can support this by prioritizing visuals, interactive elements, or concise explanations tailored to each task. St.Amant’s concept of change cues are stimuli that signal transitions in tasks. They can guide users through documentation and alert them to necessary adjustments, enhancing both clarity and engagement.
Decision-making dynamics
St.Amant breaks down the user‘s decision-making process into five key stages when interacting with documentation:
- Decision to seek documentation: Users decide to look for information.
- Decision to use: Users choose to continue or abandon the document based on initial clarity.
- Decision to do: Users initiate a task with the help of the document.
- Decision to proceed: Users progress from one step to the next, guided by cues.
- Decision to stop: Users determine when the task is complete.
Application in AI
These decisions reflect the user’s situational needs and cognitive states. The approach to sovereign AI should allow the AI to track user feedback, ensuring documentation remains aligned with evolving needs. For example, the AI can identify where users frequently stop reading, indicating potential content gaps, and refine the output to maintain a task flow.
Scripts and change points
St.Amant highlights “scripts” or predictable behavioral patterns in familiar environments as well as “change points” or specific cues that guide action shifts. These principles can be significant in a sovereign AI solution for technical writing.
Application in AI
By recognizing common scripts, the AI can adapt language and structure to match how users expect to interact with technical documents. For instance, in troubleshooting guides, the AI could create a standard script that starts with simple diagnostics and progresses toward more complex solutions.
Change points can be integrated into the AI by prompting users with cues for transitions, such as clear section headings or callouts indicating the next logical action. The integration with existing systems complements this approach as it ensures AI-driven cues align with familiar workflows, enhancing user trust and ease.
Folds and swaps for adaptive documentation
“Folds” and “swaps” are strategies for adjusting documentation patterns. A fold introduces new elements into an existing script without disrupting user flow, while a swap reconfigures existing steps to suit a new context. These concepts could make a sovereign AI solution more responsive to dynamic user needs.
Application in AI
If an AI tool detects that a new regulatory standard affects documentation, it can introduce relevant content as a fold, providing necessary information without altering the overall structure.
When a process needs to be adapted to a different environment, the AI could generate alternative documentation that “swaps” steps while retaining familiar cues. For instance, a procedure meant for on-site use could be reformatted for remote access with clear distinctions to manage user expectations.
Cognition-driven AI solutions
A fostered sovereign AI solution that integrates cognitive reflexes, scripts, and change points can drive usability, making documentation more intuitive and effective. As AI becomes an integral tool for technical writers, aligning these systems with human cognitive processes will become crucial in fostering a seamless, user-centered documentation experience.
By understanding and adapting to the ways users process information, AI solutions can support technical writers in creating content that’s both functional and engaging, meeting the high standards of modern technical documentation while remaining flexible for future needs.
References
[1] Kirk St.Amant (2023). The Cognition of Content Creation for Usable Designs (MEGAComm 2023 Presentation).
[2] Kirk St.Amant (2024). Cognition, Communication, and Commerce: A Communication-Based Approach to New Product Development and Design (MEGAComm 2024 Keynote).