Games, Modes, and the Depth of Fewer Inputs
How many inputs does a system really need?
This video explores how constrained input schemes produce surprising depth. Not by limiting what a player or a user can do, but by shifting complexity from the control surface to timing, context, and pattern recognition.
What Do We Know About the Back Button?
This video traces the concept of “back” across five distinct systems: Vannevar Bush’s Memex trails, Ted Nelson’s bidirectional links in Project Xanadu, the browser’s history stack, Git’s directed acyclic graph, and the scrollable logs of conversational AI. Each system makes a different wager on what kind of past should be available to the user. Each reveals something about why reversibility is an unsolved design problem.
When the Interface Flattens Into a Prompt
For most of computing history, we interacted with software through visible interfaces. But something fundamental is shifting. With the rise of AI agents (their skills, plugins) and language-based tools, the interface is compressing into a single text input box. This isn’t just a change in how we use computers and phones, it’s changing how we think. This video explores emerging paths in AI interface design, and why terminal-like experiences are appealing and can be dangerous.
Why Pie Menus Never Became the Default
This video explores one of those interface ideas that looks obvious in hindsight, works incredibly well in theory, yet never quite broke through into everyday use. Pie Menus: a type of radial menu, where options are arranged in a circle and selected by direction rather than vertical lists. Researchers and designers have long argued they can be faster and more efficient than traditional dropdown or pull-down menus. Yet, despite enthusiasm in academic circles and successful usage in games and specialised tools, they’re still absent from most mainstream software.
When Interfaces Decide How We Think
Were “information accidents” engineered out of our digital lives? Modern interfaces promise instant efficiency, delivering the exact answer we want, devoid of the context we need. It seems that most of the tools we use to navigate the information world today are turning “finding out” to mere consumption.
In Defence of the Spreadsheet, Reluctantly
Spreadsheets are often dismissed as legacy tools, yet they remain the most widely used interface for analysis, modeling, and decision‑making. They empower users with something unique, but the same features that give users power can also trigger catastrophes. This video tries to understand how and why spreadsheets enable both.
Verbs vs Nouns: The Word Order That Shaped User Interfaces
This video argues that the history of user interfaces is defined by a linguistic battle between two cognitive models: Noun-Verb (select object, then act) and Verb-Noun (select action, then apply to object). It posits that AI is now synthesising these opposing philosophies into a new “intent-driven” paradigm.
Interfaces of the Future Past: Can AI Revive Early Interface Concepts?
This video argues that the dominant computing paradigm, the “Xerox Consensus” of files, folders, and windows is a historical accident designed to transition 1970s office workers from paper to screens. It posits that this metaphor has imposed a “cognitive tax” on users for decades and that the rise of Artificial Intelligence (AI) and vector databases is finally allowing us to resurrect superior, associative interface concepts that were previously abandoned.
vibe coding and “wait time” brainrot
An analysis that discusses a phenomenon emerging from “vibe coding” (using AI to generate code), focusing on the cognitive risks associated with the “Wait Window”, the idle time between issuing a prompt and receiving the AI’s output.
The Curious Case of Zoomable User Interfaces and AI
Exploring the resurrection of the Zoomable User Interface (ZUI) - an infinite, continuous digital canvas - and its potential synergy with modern Generative AI. It posits that while the desktop metaphor (files, folders, windows) relies on symbolic memory, the ZUI aligns with human spatial memory, treating information as a navigable landscape rather than a stack of isolated screens.
Agents vs Operating Systems : The War That Could Kill Apps
Exploring the tension between legacy graphical user interfaces (GUIs) and the emerging paradigm of AI “agents,” arguing that while agents promise to fulfil the original, interconnected vision of computing pioneers, they risk being stifled by the same corporate forces that fragmented the web.
When Software Becomes the User: AI Agents & The Future of Interfaces
This analysis explores the emerging paradigm where artificial intelligence agents, rather than humans, become the primary operators of digital interfaces. It posits that we are witnessing a fundamental transition in human-computer interaction, driven by the ability of Large Language Models (LLMs) to interpret and manipulate graphical user interfaces (GUIs).
History of The Graphical User Interface (GUI): A Wonderful Curse
This documentary provides a critical and historical examination of the Graphical User Interface (GUI), tracing its evolution from early mechanical computation to modern mobile operating systems. It argues that while the GUI has democratised computing, its development was shaped by commercial compromises and persistent “paper-based” metaphors (such as files and folders) that may now limit human-computer symbiosis.