Key Takeaways

  • The traditional "design process proper" is dead. Andrew Ambrosino, product and engineering lead for OpenAI Codex, agrees with this, stating the old process was “predicated on the assumption that implementation is expensive.”
  • With AI, implementation is no longer the bottleneck. You can “pull all of the implementation into” the prototyping phase, making the old, exhaustive pre-build planning obsolete.
  • While the rigid steps are gone, the purpose of the design process remains vital: understanding goals, de-risking, and knowing where you are. Ambrosino calls this the "overlay."
  • Smart teams are building "baby versions" of products, like a "baby Codex" or "baby Cursor," which are dramatically simplified codebases that quickly approximate the full app's interactions for rapid testing.
  • The new design challenge shifts from meticulously planning a costly build to developing impeccable taste and curation, rapidly testing "baby versions," and making swift judgments on what to ship.

Your Design Process Is Dead. Good.

For decades, the design process felt like a sacrosanct ritual. You'd research, ideate, prototype, test, iterate, and only then carefully hand off specs for an expensive, slow implementation. It was a waterfall, or at best, a rigid agile sprint, all because building was just too costly to mess up.

Andrew Ambrosino, who leads product and engineering for OpenAI's Codex app, says that era is over. He agrees with the provocative claim that “the design process is dead,” specifically referring to “the design process proper,” as he told Lenny Rachitsky on Lenny's Podcast. Ambrosino clarifies, “that process is sort of predicated on the assumption that implementation is expensive and that you can really only afford to build once. And so you need to fully like exhaustively go through the problem space and the solution space before implementing.”

AI changes the game entirely. When AI can write and modify code on demand, the cost of "implementation" plummets. You're no longer hoarding your build budget for one big bet. Ambrosino notes, "We pulled prototyping into that. The problem now is that you can pull all of the implementation into that." The very act of building becomes part of the exploration.

The "Overlay" Still Matters: De-Risk and Discover

Before you throw out every whiteboard marker, Ambrosino offers a crucial distinction. While the specific, rigid “day-to-day specifics of the process” are gone, the underlying purpose is more important than ever. He calls this the "overlay" of the process.

“To say the design process is dead, I feel like it's both true and false, right?” Ambrosino says. “If you are if you are tied to the tools in the ex like the spec the exact like day-to-day specifics of the process then yeah it's dead like you're not going to have a good time but to throw the process out completely or throw like the overlay of the process the like hey we're at this point in the process like that is still more important than ever.”

This "overlay" is about understanding: Where are you? What are your goals? How can you de-risk your next move? These are timeless questions, now asked in a world where answers can be built and tested in hours, not months. The new role of the designer (and founder) isn't to draw perfect wireframes for a distant future, but to rapidly define, test, and refine immediate realities.

Build "Baby Versions" to Fail Faster

So, what does this new, fluid process look like in practice? Ambrosino points to "baby versions" of products. Think of it as extreme prototyping that blurs the line with actual production code.

“Many companies right now have this idea of like a baby version of the product like baby cursor,” Ambrosino explains. “You've seen this on Twitter like we have baby Codex, right? a dramatically simplified codebase that approximates all of the interactions of the production app and therefore is a lot quicker to vibe code over.”

These "baby versions" aren't just mockups; they are functional, albeit stripped-down, applications. They allow teams to "vibe code" – a term suggesting rapid, almost intuitive iteration – to explore user interactions and product feel with minimal overhead. The goal isn't perfect code or polished UI, but validated learning at light speed. The hard part isn't building anymore; it's discerning what to build, how to present it, and when to pivot. Taste and curation become your most valuable skills.

What to Do With This

Stop planning for expensive implementations that don't exist. This week, identify a core feature you're considering. Instead of a detailed spec or high-fidelity prototype, build a "baby version" – a dramatically simplified, functional codebase that captures the essential interaction. Give it to 3-5 users and gather immediate feedback, aiming to scrap or ship within 48 hours. Focus your energy on rapid judgment and iteration, not exhaustive pre-build design.