The new paradigm for building powerful tools is to design them for AI models. Instead of complex GUIs, developers should create simple, well-documented command-line interfaces (CLIs). Agents can easily understand and chain these CLIs together, exponentially increasing their capabilities far more effectively than trying to navigate a human-centric UI.

Contrary to typical design leadership, Anthropic's Head of Design advocates for minimalist interfaces like the CLI. The philosophy is that the UI is merely a medium, and the goal is to provide the purest, most direct access to the underlying technology. The focus is on the work product, not the intermediary tooling.

While GUIs were built for humans, the terminal is more "empathetic to the machine." Coding agents are more effective using CLIs because it provides a direct, scriptable, and universal way to interact with a system's tools, leveraging vast amounts of pre-trained shell command data.

Design is often mistaken for aesthetics, like choosing a border radius. Its real function is architectural: defining the simplest possible system with the fewest core concepts to achieve the most for users. Notion's success, for example, comes from being built on just blocks, pages, and databases, not from surface-level UI choices.

The terminal-first interface of Claude Code wasn't a deliberate design choice. It emerged organically from prototyping an API client in the terminal, which unexpectedly revealed the power of giving an AI model direct access to the same tools (like bash) that a developer uses.

At OpenAI, the first question is "Can we solve this with the model (tokens) instead of pixels?" This treats the AI as the primary design material, pushing designers to think about interaction and behavior before creating bespoke user interfaces.

Unlike graphical interfaces that use progressive disclosure to hide complexity, CLIs for developers demand high information density. The linear nature of terminal output means all relevant information must be presented upfront, especially when model reliability is low.

When prototyping new AI-powered ideas, build them as command-line interface (CLI) tools instead of web apps. The constrained UI of the terminal forces you to focus on the core workflow and logic, preventing distraction from visual design and enabling faster shipping of a functional version.

When designing ambiguous systems, resist creating visual mockups immediately. First, establish alignment on the fundamental concepts or "primitives." At Paradigm, this meant defining the core objects of a 'workflow' to ensure the team shared a mental model before exploring any UI.