AI and cognition
A place to gather ressources on cognitive sciences and cognitive engineering.
Core Books on Cognitive Engineering & Ergonomics:
Foundational Texts:
"Cognitive Systems Engineering" by Jens Rasmusse
Here's a living briefing document capturing our journey into cognitive engineering and ergonomics:
Our journey began 6 months and 3 weeks ago when Romain Peter established the foundation for our "Cognitive Engineering and Cognitive Ergonomics Learning." Our core objective is to build a robust understanding, integrating theory with practical application, especially concerning AI and autism management. We believe our philosophical backgrounds will allow us to bring a unique perspective to the lived experience of cognitive enhancement.
To kickstart our learning, our "Quick Start Action Items" included:
* Beginning with Don Norman's "The Design of Everyday Things."
* Joining ResearchGate and following key researchers in cognitive ergonomics.
* Systematically documenting personal experiences, which is crucial for our self-study cases.
We structured our learning into weekly breakdowns for digestible chunks, immediate practical application, and regular reflection, progressively increasing in complexity.
**Week 1: Introduction to Core Concepts**
We focused on foundational concepts, including cognitive load theory, executive function in autism, and human-computer interaction for neurodiversity. We explored how AI might intersect with cognitive processes and began identifying affordances in our daily environment. We also analyzed tools for their ergonomic design and started journaling about personal experiences with executive function, considering where AI tools could offer support.
**Week 2: Deep Dive into Norman and Behavioral Design**
We deepened our understanding by diving into Don Norman's "The Design of Everyday Things," focusing on affordances, signifiers, and "Knowledge in the Head and in the World." This directly relates to distributed cognition and offloading mental work, a concept Romain is already familiar with from David Rabouin's work. We applied these concepts by identifying poorly designed objects, evaluating our workspaces for cognitive support, and documenting instances where cognitive load visibly affected our performance or decision-making. We also introduced "Bridging the Value-Action Gap," identifying struggles with consistent AI tool usage, and analyzed habit-forming AI tools using Nir Eyal's HOOK model: Trigger, Action, Variable Reward, and Investment.
**Parallel Discoveries and Resource Compilation**
As we progressed, Romain Peter began compiling a comprehensive list of "Core Books on Cognitive Engineering & Ergonomics," "Key Academic Articles," and "Online Resources," ensuring a broad and deep understanding of the field.
A key concept we explored and assimilated was **Affordance**. We defined it as "a relationship between a physical object and an agent... that determines how the object can be used," emphasizing that "visible affordances provide strong clues about how things work." We delved into the distinction between affordances, perceived affordances, and **signifiers**, noting that "Signifiers communicate where the action should take place and are more important than affordances for designers." We extended this concept to AI tool development, highlighting its role in "Guiding Interaction and Understanding," "Facilitating Human-AI Collaboration," and "Ensuring Ethical and Responsible Use." Romain Peter specifically noted how these principles "resonate strongly with the goals of reducing cognitive load and creating predictable patterns, especially for neurodivergent users navigating the digital landscape." We also documented "Key Research Findings on Web/Desktop Tools for Neurodivergent Users," focusing on visual processing, information architecture, interaction patterns, and sensory considerations.
We also introduced the concept of **Discoverability**, aiming to assimilate its principles in cognitive engineering. We identified key aspects to explore, including "Affordances & Signifiers," "Natural Mapping," "Feedback Mechanisms," "Constraints," and "Visibility of System Status." We also prioritized "Cognitive Load Considerations" such as "Recognition vs. Recall," "Mental Models," "Information Architecture," and "Progressive Disclosure," and planned to examine "Real-world Applications" in various design contexts.
Simultaneously, we began structuring a formal course, "Cognitive Engineering and Ergonomics: Designing for the Human Mind."
* **Module 1: Introduction to Cognitive Engineering and Ergonomics** defined the field and its core principles, emphasizing user-centered design, cognitive compatibility, error tolerance, and feedback.
* **Module 2: Human Perception and Attention** covered visual and auditory perception, and different types of attention, discussing their implications for display and interface design.
* **Module 3: Human Memory and Knowledge Representation** explored working memory, long-term memory (semantic, episodic, procedural), mental models, and knowledge representation to inform design for learning and recall.
* **Module 4: Human Problem Solving and Decision Making** examined problem-solving strategies, decision-making models, and common cognitive biases (e.g., availability heuristic, confirmation bias, anchoring) and how to mitigate their impact.
* **Module 5: Human Error and System Reliability** addressed the causes and classifications of human error (slips, lapses, mistakes, violations) and strategies for prevention and mitigation, emphasizing safeguards and safety culture.
* **Module 6: Usability and User Experience Design** integrated cognitive principles into practical usability and UX design, covering principles of usability (learnability, efficiency, memorability, error prevention, satisfaction), the User-Centered Design (UCD) process, user research methods, usability testing, information architecture, interaction design, visual design, and accessibility.
* **Module 7: Real-World Applications of Cognitive Engineering** showcased applications in AI interfaces, healthcare systems, transportation, and personal productivity tools, highlighting how cognitive engineering principles are implemented.
* **Module 8: Cognitive Engineering in AI and Automation** delved deeper into AI's unique challenges and opportunities, focusing on understandable AI (XAI), human-AI collaboration, building trust in AI, and mitigating negative consequences like skill degradation.
* **Module 9: Case Studies and Emerging Trends** examined real-world examples, both successful (e.g., GUI, modern cockpit design) and unsuccessful (e.g., Three Mile Island accident, Boston's Big Dig traffic management), to provide valuable lessons and identify factors for success or failure. This module also explored emerging trends such as physiological computing, adaptive interfaces, and cognitive engineering for an aging population.
* **Module 10: Applying Cognitive Engineering to Personal and Professional Development** shifted focus to enhancing our own cognitive performance, covering strategies for learning, decision-making, productivity, and, notably, applying principles to autism management.
This structured approach, combining theoretical learning with practical application and ongoing documentation, ensures we are building a robust and deeply understood foundation in cognitive engineering and ergonomics.By Romain Peter