Empirical Bottoms-Up Development
by Alexander Embiricos • Product Lead for Codex at OpenAI
Former startup founder (screen sharing/pair programming), former PM at Dropbox. Now leads the product team for OpenAI's coding agent, Codex.
🎙️ Episode Context
Alexander Embiricos discusses the evolution of Codex from a code completion tool to a proactive software engineering 'teammate.' He explores OpenAI's unique 'empirical bottoms-up' product culture, the massive acceleration of internal development (building Sora's app in 28 days), and the future of agentic workflows where AI proactively acts on team chatter and signals.
Problem It Solves
Prevents over-engineering features that don't work with stochastic models and reduces the 'time to reality' for novel AI capabilities.
Framework Overview
A product development philosophy that prioritizes rapid prototyping ('vibe coding') and internal dogfooding over rigid long-term planning. It acknowledges that in the AI era, technological capabilities change too fast for traditional 'aim then fire' roadmaps.
🔄 Iterative Cycle
When to Use
When building on top of rapidly evolving infrastructure (like LLMs) where user behavior is emergent and unpredictable.
Common Mistakes
Spending too much time on 'Ready, Aim' (strategy decks) and not enough on 'Fire' (shipping and testing), or assuming traditional PM frameworks apply to AI.
Real World Example
The Sora Android app was built in 18 days by a tiny team using Codex to port logic from iOS, launching to the public 10 days later.
We can have really good conversations about what's happening in low months or weeks... But there's this awkward middle ground... it's much more important for us to be very humble and learn a lot more empirically.
— Alexander Embiricos