Andrew Ambrosino
Product and engineering lead for OpenAI Codex; frames agent-era product work around cheap implementation, taste, role overlap, and delegated work surfaces.
Andrew Ambrosino is a product and engineering lead for the OpenAI Codex app. In a Lenny's Podcast interview, he describes Codex as moving from a developer tool toward a broader work surface where people can start, track, automate, and coordinate work across applications.1
His useful lens is cross-functional: he has worked across design, engineering, product management, and founding companies, and uses that background to argue that AI-Native Product Work makes role boundaries more porous without eliminating the underlying disciplines. Product, design, engineering, finance, and other functions still have knowable skills and best practices even when AI lowers the tool barrier.1
Ambrosino's recurring theme is that when implementation becomes cheap, taste and curation become the bottleneck. Teams can generate many prototypes or agent-built artifacts, but someone still has to choose the right medium, decide what the artifact proves, preserve product coherence, and know when a feature is ready.1
In the blog
SummaryOpenAI Codex and the new shape of product work
Mentioned in 6 pages
Engineering with agents
Work, teams and companies
People & organisations