1.1 HCI Lecture Notes (Transcript)
What is Human-Computer Interaction (HCI)?
HCI is the study and design of interactions between humans and computers, encompassing research, evaluation, and the design of interfaces.
The field considers every element of the human: perception (senses like seeing, hearing, feeling; and to a lesser extent smell and taste), memories, experiences, skills, and knowledge, because these shape how we interact with technology.
The core idea is to make the interface disappear into the task so the user spends more time thinking about the task and less about the interface.
This course emphasizes that humans and computers together interact with a task, mediated by the computer, rather than a simple one-way hand-off of interaction.
Real-world ubiquity: computing is everywhere, so HCI is everywhere (mobile devices, AR, etc.). AR games like Pokémon GO illustrate how the entire world can become a context for HCI.
The Human-Computer Interaction Triad and Task Mediation
Two primary views of interaction:
Traditional view: the human interacts with the computer, and the computer responds to the human.
Task-centric view: the human interacts with the task through the computer; the computer mediates the interaction.
The ideal goal: make the interface as invisible as possible so the user focuses on the task, not the interface.
Practical reality: interfaces often remain visible, but designers should minimize the cognitive load and friction to let users think about the task.
Disappearing interfaces: aim to design tasks so the interface vanishes-in-the-middle of interaction.
Examples of interface-invisibility in practice:
Video games can feel like the player is inside the game world due to intuitive controller design and learned mappings (e.g., forward moves forward, backward moves backward).
Overly multiple remotes can keep the user focused on how to operate the interface rather than the task (e.g., watching TV vs controlling the same system with two remotes).
The Field’s Ubiquity and Responsibilities
HCI is everywhere as devices become more capable and ubiquitous (cars, wearables, home devices, etc.).
Cars like Tesla exemplify computer-on-wheels: dashboards and almost all controls are computerized.
The field blends engineering, psychology, design, and cognitive science to study how people perceive, think, and interact with technology.
The danger of overconfidence: being good at using computers does not guarantee expertise in designing interactions for others.
Knowledge vs. unknowns: a common teaching metaphor shows the progression from thinking you know a lot to realizing there is much more to know as you learn more about HCI.
History and Evolution: UI/UX vs HCI, and Foundational Concepts
Early UI innovations focused on screen-based interaction: light pens, the computer mouse, etc.
Foundational UI design principles include:
Grids and typography for organizing content
Menu design and platform comparisons (e.g., Mac vs PC menu placement)
Responsive design and adapting interfaces to different screen sizes
Prototyping techniques using pen-and-paper or wireframes
UI design became a well-defined field focused on on-screen interaction; HCI broadens beyond the screen to encompass general interaction principles.
Relationship with UX design:
HCI is about understanding human interactions with computers.
UX design is about dictating those interactions to create good experiences; many view UX as a subfield of HCI.
The two are symbiotic: research informs design, and design outcomes inform further research.
Feedback cycles are central: users interact with interfaces, outcomes are evaluated, and those results feed back into both design and research.
HCI and psychology are deeply connected: research informs design, and design outcomes inform psychology’s understanding of human cognition and perception.
Georgia Tech example: HCI is cross-listed as computer science and psychology, illustrating its interdisciplinary nature.
HCI in the Broader Field and Hierarchy
HCI as a subset of Human Factors Engineering: Human factors designs interactions between people and products or systems, not limited to computing.
Subdisciplines within HCI may overlap with UI design and UX design, but the representation here emphasizes HCI as a broader framework.
As computing becomes embedded in more devices, the boundary between HCI and human factors engineering shrinks; the applications extend to everyday devices like showers, refrigerators, wristwatches, etc.
The field’s expansion means many domains now count as HCI applications (e.g., cars, home devices).
UI, UX, and HCI: Distinctions and Synergies
UI design historically focused on on-screen elements and interaction resolution; many concepts from UI design originate in HCI.
UX design focuses on dictating interactions; understanding user needs and interactions is essential for strong UX.
HCI provides the general methods and principles; UX applies them to create meaningful experiences.
The relationship is symbiotic: research informs design, design outcomes inform research, and both guide improvements in understanding human-computer interaction.
Design-based research: using the results of designs to conduct research and improve understanding.
Research, Design, and Feedback in HCI
HCI involves both research (studying users, needs, cognition, evaluation) and design (prototyping, implementing interfaces).
The design process is iterative: research informs designs, and designs are evaluated to refine research questions and methods.
Feedback cycles thread through everything: user participation, task performance, evaluation, and redesign.
HCI’s scope includes a broad range of interfaces—from traditional screens to touch, gesture, voice, VR/AR, and emerging modalities.
The principles apply across domains: from desktops to smartphones, wearables, and AR/VR experiences.
The Role of AI in HCI: Speed, Interface, and Interaction Paradigms
AI can accelerate HCI work: simulated need-finding, faster prototyping, quicker evaluations, and handling larger AB experiments.
AI is itself a computer; when AI agents interact with humans, it is a form of HCI. This motivates the study of Human-AI Interaction (a domain that could warrant its own course).
A key perspective: AI tools act as new interfaces to underlying tasks, enabling more natural interaction (e.g., natural language interfaces).
Historical trajectory: early interfaces required learning specific software; AI tools today often let users tell the system what they want rather than how to do it (Nielsen’s view of AI as a third UI paradigm).
Examples of AI in action:
Image editing: specifying “remove this person from the background” instead of hand-selecting pixels.
Code generation: suggesting or generating code rather than manually writing it from scratch.
Practical implication: AI is strong in certain areas but not all; existing design principles still apply to AI-driven interfaces.
The quote often attributed to Jacob Nielsen: AI represents a shift toward telling the computer what you want, not how to do it; this is the basis for the third UI paradigm.
Major Takeaways and Practical Implications
Design goal: make interactions feel seamless and task-focused by minimizing interface interference.
The field spans research and design; outcomes in one inform the other in a continuous loop.
HCI is deeply interdisciplinary, merging engineering, psychology, design, and cognitive science to understand perception, cognition, and interaction.
Ubiquity means every new device capable of computing can be a subject of HCI, expanding both opportunities and responsibilities (usability, accessibility, privacy).
AI represents both a tool to accelerate HCI work and a new class of interface that changes how users interact with tasks; it demands careful design to ensure reliability, transparency, and user control.
Design-based research and feedback cycles enable iterative improvement and grounding in real user data.
Ethical and practical implications to consider explicitly:
Accessibility and universal design: ensuring interfaces work for diverse users and contexts.
Privacy and data security: pervasive computing and AI raise concerns about data collection and control.
Transparency and user agency: AI-driven interfaces should be explainable and allow user override when appropriate.
Bias and fairness: AI systems can reflect or amplify biases; designing inclusive interfaces helps mitigate harm.
Cognitive load and fatigue: even invisible interfaces can impose mental effort; balancing speed with comprehension is key.
Connections to Previous Lectures and Real-World Relevance
HCI sits at the intersection of engineering and psychology, reflecting how mental models, perception, memory, and cognition inform design choices.
Real-world case: Apple’s 1992 workspace study informed an interface that reflected how people organize information, linking psych insight to practical design.
The broader trend of computing being integrated into everyday objects (cars, home devices) illustrates the shrinking boundary between traditional UI design and broader human factors engineering.
The AI discussion connects to ongoing industry shifts toward natural-language interfaces and “third UI paradigm” thinking, highlighting a practical trajectory for future work in HCI.
Key Terms and Concepts (glossary style)
Disappearing interface: design goal to minimize users’ focus on the interface so they can concentrate on the task.
Task-mediated interaction: the computer acts as an intermediary between the human and the task.
Feedback cycle: iterative loop of research → design → evaluation → redesign.
Design-based research: research approach in which design artifacts are used to generate knowledge.
Universal design / accessibility: designing for all users, regardless of ability.
Human factors engineering: broader field integrating psychology and engineering to optimize human-system interaction, not limited to computing.
User experience (UX) design: designing how users experience interactions with products and interfaces.
User interface (UI) design: traditional focus on screen-based, tangible controls and displays.
Human-AI Interaction (HKI or HAI): intersection of HCI with artificial intelligence, focusing on how humans interact with AI systems.
Mathematical Notes (conceptual, with LaTeX)
Interface burden and task efficiency relationship (conceptual):
The efficiency of task completion E is inversely related to interface burden B, i.e., E \propto \frac{1}{B}
If we model a simple feedback loop where user understanding U improves with iterations k of design and evaluation, one could denote:
U{k+1} = Uk + \alpha \, ext{(quality of feedback)} where (0 < \alpha \le 1) and the feedback quality increases with rigorous evaluation.
AI as interface paradigm can be framed as a new interaction mode; denote three primary modalities as: traditional (M1), touch/gesture (M2), and natural language AI-driven (M3). A simple comparative utility model might be:
U(M1) < U(M3) \text{in natural-language-driven tasks}
but with caveats for reliability, interpretability, and domain suitability.
Quick References for Study
Core ideas to remember:
HCI encompasses more than screen-based UI; it’s about human-centered interaction across contexts and devices.
The goal of invisibility of interface is a guiding principle, but not always fully achievable.
The field is iterative and interdisciplinary, blending design and research.
AI changes the interaction paradigm by enabling users to tell computers what they want, not just how to do it, but it introduces new design challenges.