In recent years, tech giants and startups alike have attempted to replace traditional screens, buttons, and menus with conversational AI. From the Humane AI Pin and Rabbit R1 to the integration of Meta AI and Copilot buttons, the industry has pushed to eliminate graphical user interfaces (GUIs). Yet, many of these efforts have failed. This is because traditional GUIs excel at simple, habitual tasks. Opening an app to order coffee or call an Uber takes seconds via physical tapping, whereas asking a voice assistant or chatbot can be slower and more frustrating.
The Rise of Agentic Tools and Headless Tech
On the other end of the spectrum, highly complex tasks are seeing an organic revolution. Developers and everyday users are adopting command-line interfaces (CLIs), Model Context Protocols (MCPs), and developer agents like Claude Code to automate tedious, multi-step workflows—such as filing taxes, organizing files, and generating custom animations. In response, modern software companies are adapting. Products like Linear and PostHog are simplifying their interfaces, while Salesforce, Google, and Notion are offering headless versions and CLIs meant to be operated by AI agents rather than human eyes.
Direct vs. Indirect Manipulation
To understand the future of human-computer interaction, we must understand why GUIs worked in the first place: direct manipulation. Tapping, pinching, and dragging allow users to see all available options upfront. However, as software grows in complexity, direct manipulation breaks down, leading to intimidating, cluttered interfaces like Adobe After Effects or Blender. Conversely, chatbots use indirect manipulation—allowing users to bypass complex menus through free-form text. While simple tasks favor GUIs and highly complex tasks favor AI agents, the vast majority of daily tasks lie in the middle. This middle ground is where the next technological shift is happening.
The Solution: Generative UI and Its Three Levels
The bridge between simple GUIs and complex chatbots is Generative UI—interfaces created on the fly by AI based on user intent. This concept evolves through three distinct levels:
- Level 1: Dynamic Content Generation. Instead of returning a wall of text, the AI generates a clean, interactive graphical widget (e.g., Claude rendering a custom interactive visualization of a complex concept).
- Level 2: Hybrid Adaptive UI. The underlying structure remains a standard app, but specific components adapt to highly unique, complex queries. This changes product design fundamentally; instead of designing static screens, designers build “rules” and “UI LEGO blocks” that the AI assembles dynamically.
- Level 3: Fully Generative Products. The entire interface of the application is created dynamically on the spot, designed specifically for the unique context of the user’s prompt.
Conclusion: The Hybrid Future of Tech
The future of technology will not be a single gray chatbot rectangle, nor will it be the static apps of the last 15 years. Instead, it will be a hybrid ecosystem. Simple tasks will rely on traditional GUIs with direct manipulation; repetitive, backend workflows will be entirely automated by background agents; and complex, personalized tasks in the middle will utilize adaptive Generative UI.
Mentoring question
How can you leverage the hybrid approach of traditional user interface, agentic automation, and generative UI to improve user experience or productivity in your current projects?
Source: https://youtube.com/watch?v=f32W5BEzWN0&is=M9zrm9tIPif-QURV