Forcing users to choose specific models, adjust reasoning parameters, or configure agent workflows creates severe configuration fatigue. Sottiaux argues that product interfaces should eventually disappear into a simple, ambient interaction where users only specify communication channels and goals, letting underlying models handle execution automatically.

In an agentic world, the core AI model becomes a commodity. The defensible product is the curated experience layer built on top of it—the guardrails, instructions, and personality that define the user interaction and differentiate the offering.

The user interface is becoming invisible as AI models become the primary content consumption layer. Product teams must now focus on how their content is used within these models, measuring value through API calls and attribution in AI-generated outputs, not website clicks or session times.

The rise of AI means product development is shifting from a UI-centric model to a programmatic-first approach. As agents become the primary "users," the focus moves to APIs and other machine-to-machine interfaces. The traditional graphical user interface is becoming a secondary concern, mainly for human oversight.

A huge portion of product development involves creating user interfaces for backend databases. AI-powered inference engines will allow users to state complex goals in natural language, bypassing the need for traditional UIs and fundamentally changing software development.

AI agents make software UIs secondary. The real product is now the headless infrastructure layer of data and APIs that agents interact with. This means companies must shift their primary focus from pixel-perfect interfaces to building robust, accessible, and well-documented infrastructure.

Building reliable AI products requires designing two interfaces: one for the user (UI) and one for the AI agent. This "agent interface" is a library of tools, APIs, and primitives that constrains the AI. It provides deterministic grounding, preventing hallucinations and ensuring the AI operates accurately and efficiently.

Building text or voice-first AI agents means discarding years of UI-based product development principles. Without the "crutch" of a visual interface, product managers must solve novel challenges in reliability, user education for new mental models, and creating a magical experience through conversation alone.

Designing for AI is less about crafting pixel-perfect UIs in Figma and more about creating the underlying system or "harness." This involves enabling the agent to perform long-running tasks, verify its own work, and operate effectively within technical constraints, which is where the real design work lies.