The Evolution of UX

How did we get from Line commands to AI agents?

How interface design shifted agency from users to systems through four key phases:

𝟭. 𝗠𝗮𝗻𝘂𝗮𝗹 𝗨𝗫 (𝟭𝟵𝟳𝟬𝘀-𝟭𝟵𝟴𝟬𝘀) 🛠️

• User agency: High (users control every action)

• Key innovation: Text-based interaction ("Mother of All Demos" by Douglas Engelbart)

• Example: UNIX terminals (1970s) required memorizing commands like "rm -rf" for file deletion and "grep" for pattern matching

𝟮. 𝗚𝘂𝗶𝗱𝗲𝗱 𝗨𝗫 (𝟭𝟵𝟴𝟬𝘀-𝗣𝗿𝗲𝘀𝗲𝗻𝘁) 🖥️

• User agency: Moderate (systems suggest paths)

• Key innovation: GUI, Visual metaphors (Xerox Parc, Mark Weiser)

• Example 1: Macintosh Finder (1984) introduced drag-and-drop file management and trash can deletion .

• Example 2: ATM interfaces (1980s) with step-by-step withdrawal/deposit workflows

𝟯. 𝗔𝘀𝘀𝗶𝘀𝘁𝗲𝗱 𝗨𝗫 (𝟮𝟬𝟭𝟬𝘀 - 𝗣𝗿𝗲𝘀𝗲𝗻𝘁) 🤖

- User agency: Shared (systems handle specific tasks)

- Key innovation: Predictive help

- Example 1: Google Autocomplete (2008) anticipates search queries after 2-3 keystrokes

- Example 2: Siri (2011) enabled natural language requests for calendar management

𝟰. 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗨𝗫 (𝟮𝟬𝟮𝟯-𝗣𝗿𝗲𝘀𝗲𝗻𝘁) 🤖 🧠

As @Kwame Nyanning describes it "Agentic UX is the overall experience a person has with an autonomous agent or agent-like product, system, or service. Agentic interfaces are composed in real-time by the agent within a generative canvas"

- User agency: Delegated (systems achieve goals independently)

- Key innovation: Autonomous problem-solving

- Example 1: Devin AI (2024) software engineer that writes/tested code through iterative debugging

- Example 2: ChatGPT Plugins (2023) coordinate flight bookings by comparing prices/availability

🎯𝗪𝗵𝘆 𝗧𝗵𝗶𝘀 𝗠𝗮𝘁𝘁𝗲𝗿𝘀?

We're entering an era where UX designers become "𝗴𝗼𝗮𝗹 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘀" rather than interface sculptors. The next frontier? Designing responsible agent ecosystems that balance automation with user oversight.

𝗞𝗲𝘆 𝗣𝗮𝘁𝘁𝗲𝗿𝗻𝘀 𝗜𝗻𝗵𝗲𝗿𝗶𝘁𝗲𝗱 𝗳𝗿𝗼𝗺 𝗣𝗿𝗲𝘃𝗶𝗼𝘂𝘀 𝗣𝗵𝗮𝘀𝗲𝘀

👤 𝗨𝘀𝗲𝗿-𝗖𝗲𝗻𝘁𝗲𝗿𝗲𝗱 𝗗𝗲𝘀𝗶𝗴𝗻: Agentic UX inherits the focus on user needs from Guided UX, ensuring that agents are designed to serve user goals effectively.

⚙️ 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗼𝗻 𝗮𝗻𝗱 𝗦𝗵𝗮𝗿𝗲𝗱 𝗔𝗴𝗲𝗻𝗰𝘆: The automation and assistance patterns from Assisted UX are amplified in Agentic UX, where agents take on more autonomous roles.

🔄 𝗔𝗱𝗮𝗽𝘁𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝗮𝗻𝗱 𝗣𝗲𝗿𝘀𝗼𝗻𝗮𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻: The adaptability seen in Assisted UX is further enhanced in Agentic UX, with agents capable of multi-modal interactions and personalized narratives.

𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗨𝗫 is NOT just a new frontier! It is a culmination of lessons learned from its predecessors.