Designing the Future of Human<>Agent Collaboration

What if you could test a hundred times more ideas before writing production code? AI agents are making it possible.

#1about 4 minutes

Figma's evolution into a product development platform

Figma has grown from a design tool to a full platform where developers are a primary audience, closing the loop from idea to shipped product.

#2about 2 minutes

The core beliefs behind building great products

The best products are built through extensive iteration and by collaborative teams that include diverse roles from the very beginning.

#3about 2 minutes

Understanding AI agents as collaborative assistants

AI agents should be viewed as active collaborators and assistants that augment human skills like critical thinking, not as replacements for team members.

#4about 5 minutes

Using AI for rapid ideation with Figma Make

Figma Make allows teams to go from a simple prompt to a functional, code-backed prototype, enabling the exploration of hundreds of ideas quickly.

#5about 8 minutes

Bridging prototypes and production with Figma MCP

The Figma MCP server provides design intent and context to AI coding assistants, enabling them to generate production-ready code that uses your existing codebase's abstractions.

#6about 2 minutes

How the MCP server provides context to LLMs

The MCP server feeds LLMs crucial context including design system patterns, component references, image data, and pseudo-code to generate accurate UI.

#7about 3 minutes

A walkthrough of the design-to-code workflow

A demonstration shows how to copy a link from a Figma selection and use it within an AI-powered IDE like Cursor to implement the design directly into the codebase.

#8about 3 minutes

The developer's evolving role in the age of AI

AI tools shift the developer's role from pure implementation to problem-solving and orchestrating agents, using the time saved to iterate more and build better products.