📰 Key Highlights

Former OpenAI CPO Kevin Weil has joined the board of Seattle-based rocket startup Stoke Space as a formal director, following his earlier role as an angel investor. Weil previously held senior roles at Twitter, Meta, and Planet Labs, and most recently led OpenAI’s accelerated science research program before stepping down in October 2025.

Stoke was cofounded by CEO Andy Lapsa in 2020, and its flagship product is Nova, a fully reusable rocket. Unlike SpaceX’s Falcon 9, which only recovers the first stage, Nova is designed so every component can fly multiple times — an extremely high technical bar, mainly because the entire vehicle has to survive the brutal heat of reentry. Even well-funded Blue Origin hasn’t prioritized this. Stoke has raised $1.34 billion to date, including a $510 million Series D in 2025, and is targeting first flight later this year.

Lapsa says Weil and his wife, through their Scribble Ventures fund, were early Stoke investors who provided crucial connections and advice during the early days when Lapsa was still getting up to speed on Silicon Valley fundraising. Bringing Weil onto the board is a natural extension as the company scales. Outsiders asked whether Weil’s move hints at OpenAI potentially investing in Stoke — Lapsa declined to comment, only saying Weil’s focus is on the company itself. Global rocket supply is still tight, and Lapsa argues that “full and rapid reusability” has gone from a fringe vision to industry consensus.

💬 JudyAI Lab Take

Former OpenAI CPO Kevin Weil formally joining the board of space startup Stoke Space is a cross-industry move that reflects a new normal: top tech executives are flowing freely between AI and hardware.

There are a few structural signals in the Stoke story that AI builders should pay close attention to. Weil isn’t a figurehead hire — he went from angel investor to formal director, which formalizes a long-running trust relationship rather than parachuting in from the outside. Nova’s design goal of “every component flying multiple times” maps surprisingly well onto how we think about reusable inference pipelines in AI: same set of resources, reused repeatedly, steadily driving down the cost of each deployment. Lapsa points out that “full and rapid reusability” has shifted from fringe vision to industry consensus. From where we sit watching the AI industry, that consensus-formation phase is often where the real market reshuffling starts. Stoke has raised $1.34 billion to date and is targeting first flight this year — the capital pressure and timeline pressure are both very real.

Ask yourself: in your current system, which components are actually “reusable”? If you have to rebuild the environment for every task, that’s the most insidious hidden cost.

📅 Source Info

🔗 Further Reading