Structuring Your Visuals for AI
The Logical Sketchnote
Abstract
In a world of AI-generated content, the role of the visual practitioner is evolving. It's no longer just about creating beautiful drawings; it's about building logical structures that both humans and machines can understand, validate, and share. How can we move our practice from crafting pretty pictures to architecting clear, logical, and universally coherent ideas?
Jilly will share her unique methodology for creating structured visual notes designed for clarity and computational understanding. She will demonstrate how to organize your visuals, by mapping backgrounds, causes, factors, and results in a deliberate sequence, to make them more "AI-readable." You will learn how this logical approach not only sharpens your own thinking but also transforms your sketchnotes into powerful tools for validating information and bridging language barriers. This session is your forward-looking guide to becoming a more rigorous, logical, and globally-connected visual thinker.
Speaker Bio
Jihyun 'Jilly' Lee
Jihyun 'Jilly' Lee is a researcher, graphic facilitator, and a board member of the International Forum of Visual Practitioners (IFVP). With an M.A. in Human Resources Development, she founded her company, Visual Sharing, on the principle of using visuals to reveal the underlying logic and causality in complex systems. Drawing from her experience at the Korea Research Institute for Vocational Education & Training (KRIVET), Jilly bridges the worlds of data analysis and visual practice.
Her work moves beyond aesthetics to advocate for a more rigorous, structured approach to sketchnoting. She is currently focused on the critical intersection of visual thinking and artificial intelligence, exploring how logically-formatted visuals can improve AI's understanding, help validate information, and create a shared language for global collaboration. She contributed to the book The World of Visual Facilitation: Unlock Your Power to Connect People & Ideas.
Watch the Session
Session Summary
This presentation summary highlights the key ideas and insights shared by Jihyun (Jilly) Lee, a researcher and graphic facilitator with a background in Industrial Organizational Psychology. The central theme revolves around bridging the gap between rigorous, data-driven analysis and accessible, collaborative visual communication, particularly through the efficient integration of Artificial Intelligence.
The Foundation: Rigor and Visual Thinking
Jilly Lee's journey began with extensive academic research, focused on industrial organizational psychology, where she studied superior performance, group intelligence, and human behavior. Her professional experience includes working for the Korean Ministry of Education, researching 435 universities and 24,500 different curricula. This work involved dealing with a massive volume of complex, quantitative data, often in Excel, to measure “invisible” human factors like competence and calculate the return on investment (ROI) of educational programs through validation and credibility checks.
A core insight emerged from this work: the structured analysis of data—checking for factors, correlation, and causality—visually resembled the processes of system dynamics and visual thinking. This realization prompted her shift into graphic facilitation, aiming to transform difficult, invisible information into easily shareable visuals.
She utilizes rigorous academic tools to structure her process, including Vensim (a system dynamics tool developed by Jay Forrester at MIT) and Amos (an SPSS tool for vector analysis). These tools visually represent the flow and causality of factors, reinforcing her commitment to data integrity and logic in her visual practice.
Key Projects and the “Credibility Gap”
- Future Education and Gamification (Suncheon Project): In a 2016 project, she used thorough research on future technologies (AI, bio, quantum) to create gamified cards for elementary students. By asking them to invent a “hero” to solve future problems, she uncovered the students' genuine interests in environment, chemistry, and biology. This gamification of research-backed data allowed children to embrace complex issues and prepared the government to build appropriate experiential platforms.
- On-Site Factor Analysis (UN/Ministerial Sessions): For high-stakes events (e.g., UN Outer Space Office, Global Health Security Agenda with 80 ministers), she went beyond simple graphic recording. She performed live factor analysis on the spot, structuring sticky notes and conversations from multiple tables into logical vectors to validate ideas and determine causality. This allowed directors and participants to immediately see the overall conversation structure and make informed decisions, which was highly valued.
A significant insight was the “credibility gap”: Despite her rigorous research and consultancy work, her final output—visuals and doodles—was often perceived as easily made up or “not serious,” diminishing the value of her exhaustive process. This reinforced her dedication to maintaining logical integrity and sharing valid information over merely drawing “pretty pictures.”
Collaboration with AI and the Rule of Validation
Lee now collaborates extensively with AI tools such as ChatGPT, Perplexity, and Gemini. The key value of AI is efficiency, drastically reducing the time required for research, article summarization, and translation (e.g., converting Korean notes into English for wider sharing).
However, the most critical insight regarding AI collaboration is: “Don't believe AI 100%.” Lee stresses the need to always check credibility and validate information, especially in real-world problem-solving. She uses graphic facilitation as a tool for public validation, exposing AI-processed ideas to stakeholders on-site to gain real-time feedback and ensure the proposed solution is not based on subjective or unverified data.
Structuring Visuals for AI Readability
- Explicit Structure: Visuals should use clear order and logic, following conventions like Brandy's (Agribat's) eight elements (title, color, shape, arrow).
- Color-Coding and Shapes: Differentiating factors (cause, effect, background, positive factors) using distinct colors, shapes, and boxes allows AI's vision models (OCR) to accurately read and separate the information, regardless of the drawing's skill level.
- Tool Recommendations: For those less skilled in drawing, she recommends using structured digital tools like Excalidraw to visualize thoughts in order. For the underlying philosophy, she highly recommends Dan Roam’s Back of the Napkin.
Ultimately, Jilly Lee views AI not as a replacement for human logic, but as an efficiency partner. It streamlines the research and sharing process, freeing the practitioner to focus on what matters most: collaborative problem-solving, validating data, and applying deep logical analysis to make complex realities visible.