Y Combinator Startup Podcast
Y Combinator Startup Podcast

The End of the Designer–Engineer Divide

Head of Design Ryo Lu helped transform Cursor from a feature-layer on top of VS Code into one of the world's leading AI code editors.He joins YC's Aaron Epstein on Design Review to talk about the path that brought him to Cursor, how rapid prototyping reshaped the core product and how he&#x

Featured Speakers

Y Combinator HostRyo Liu Guest

Topics Discussed

Episode Summary

Executive Summary: Ryo Liu, head of design at Cursor and former designer at Notion/Stripe/Asana, argues that AI is collapsing the gap between design and engineering. He says designers should become builders, use agents to prototype directly in code, and focus on core systems and craft rather than static mockups. Cursor’s product and brand evolved around simplifying interfaces, defaulting to agents, and enabling rapid iterative “sculpting” of real software.

Main Topics: From self-taught builder to design leader (Priority: 4/5): Ryo describes starting by making websites at 11, learning by building rather than through formal design/engineering boundaries, and later moving through startups, Asana, Stripe, and Notion before Cursor. AI blurs roles and makes designers coders (Priority: 5/5): He argues that designers can now build directly with agents, learn by doing, and use code as a shared language with engineers, reducing the historical separation between design, product, and engineering. Cursor’s design philosophy: unify and simplify (Priority: 5/5): Cursor’s biggest product change was merging chat, composer, and agent into one agent-first experience, replacing scattered UI states with simpler defaults and layered power tools. Systems-first design over feature accumulation (Priority: 5/5): Ryo contrasts traditional feature-by-feature design with a systems-first approach built on stable primitives; he says products scale best when core concepts stay simple and configurable. Design workflow in the AI era: prototype in real code (Priority: 4/5): He showcases Baby Cursor and RiOS as examples of fast, code-based exploration where designers can build living prototypes, test interactions, and iterate without waiting for engineering. Future of interfaces: adaptive, decomposed, and context-aware (Priority: 4/5): Ryo predicts interfaces will persist but be decomposed and recomposed by AI based on context, user preference, and task, while still preserving familiar visual structures like tables, to-do lists, and previews. Advice for founders and designers (Priority: 3/5): His advice is to identify the core concepts that define a product over 10 years, then build from those primitives and let AI accelerate iteration and evolution.

Key Arguments: Designers should stop treating code as someone else’s domain; AI tools now let them build real products directly and learn through iteration. Cursor’s product became clearer and more powerful when chat, composer, and agent were unified into a single agent-first workflow. The best products scale through stable primitives and systems thinking, not by endlessly adding features and buttons. AI doesn’t remove craft; it shifts craft toward sculpting, refining, and making imperfect outputs excellent. Real prototypes are better than static mockups because they expose interaction, motion, and state, which are now easier to explore in live code. Interfaces won’t disappear, but AI will increasingly recompose them around the user, task, and context. Designers and engineers will increasingly share code as a common language, blurring historical role boundaries.

Data Points: Cursor user base: more than 1 million people worldwide - Cursor is described as the leading AI coding tool used globally. Cursor company size at join: 20 people - Ryo says Cursor had about 20 people when he joined. Cursor current company size: 250 people - He notes the company has grown substantially since he joined. Design team size: 4 people - Ryo says the design team remains very small, including himself and three others. Time at Cursor: almost 10 months - He describes his tenure as feeling like “five years” in AI time. Product speed improvement: 4x faster than Sonnet - He says Cursor’s new composer model is four times faster than Sonnet. Product cost improvement: 2x cheaper than Sonnet - He says the new composer model is also twice as cheap. Age when first making websites: 11 - Ryo says he started making websites when he was 11. Age when making the Apple fanboys site: 17 - He says a popular fan site in China was made when he was 17.

Pivotal Quotes: "My personal KPI at Cursor this year is to turn all the designers into coders." — Ryo Liu: Opening thesis on role-blurring and empowering designers to build. "The roles will start blurring, the designers will start coding, the engineers will start designing, and then our shared language is code." — Ryo Liu: Explaining the future of cross-functional product building. "If you get something bad or like ugly, it's actually your job to make it pretty the way you want it." — Ryo Liu: Describing the new craft of sculpting AI-generated outputs.

Implications: Design teams should adopt AI-native workflows, learn basic coding, and focus on systems and interaction quality. Products will increasingly be built by mixed-discipline teams using code as the common medium, with AI accelerating iteration and personalization.

🔓 Sign Up for Unlimited Episode Search

About Y Combinator Startup Podcast

We help founders make something people want. The Y Combinator Podcast is where builders talk about building. From the earliest days of an idea to scaling a company that changes the world, YC partners and founders share real stories, lessons, and tactics from the frontlines.

View all episodes from Y Combinator Startup Podcast