Why not deliver a design agent that can guide the entire process or group of processes vs just a single tool?


That's actually the direction things are heading. As AI products have evolved, we've moved from single-purpose tools to agentic systems where AI models have access to multiple tools, can plan which ones to use and how, and execute multi-step processes. This is exactly what you're describing - an AI that guides an entire workflow rather than just completing one task.

The challenge is that agent management becomes a critical part of these products' design. How can people start, steer, and stop multiple agents (and subagents) and stay on top of their results? When you have agents running in the background coordinating multiple processes, the interface needs to support monitoring and managing lots of agentic workflows concurrently instead of guiding one at a time.

We're already seeing this play out. When agents can use multiple tools, call other agents and run in the background, a person's role moves to kicking things off, clarifying things when needed, and making use of the final output. The chat interface becomes less prominent because you're not constantly going back and forth - you're reviewing results when they're ready.

The next frontier is agent to agent collaboration. Google's recently announced Agent to Agent protocol is specifically designed to enable "multi-agent ecosystem across siloed data systems and applications." Does this result in very different product experiences? Probably. What does it look like exactly? I don't know yet, but it's coming fast.

That said, even with sophisticated agents, a design needs a voice. It needs a clear message. It needs a personality. The human role of providing vision and unifying diverse aspects into a cohesive message remains essential - agents just help execute that vision more effectively.