Kevin Weil, the former OpenAI chief product officer, has joined the board of Stoke Space, the Seattle startup that has raised $1.34 billion to build a fully reusable rocket. Weil takes a director seat at a company whose $510 million Series D last year vaulted it into the small club of well-capitalized SpaceX challengers. The move ties one of the most closely watched product executives in AI to a launch-vehicle bet that will hinge on a first flight targeted for this year.
Weil's résumé runs through Twitter, Meta, Planet Labs and OpenAI, but his link to Stoke is older than any of those tours. He and his wife Elizabeth Weil were early investors in the company through their fund Scribble Ventures, backing Stoke shortly after Andy Lapsa cofounded it in 2020 and enrolled in Y Combinator's winter batch.
Lapsa, a former Blue Origin engineer, framed Weil's value as the network and fundraising instincts he lacked coming out of hard tech.
Key facts
- 01Kevin Weil, former OpenAI chief product officer, joined Stoke Space's board as a director.
- 02Stoke Space has raised $1.34 billion total, including a $510 million Series D in 2025.
- 03Weil served as OpenAI CPO from June 2024 to October 2025, then led scientific research until April.
- 04Sam Altman was reportedly exploring an investment in Stoke last year as a possible SpaceX competitor.
- 05Stoke is building Nova, a fully reusable rocket targeting a first flight this year.
Weil left OpenAI's scientific-research effort in April after that program was distributed across the frontier lab. He had served as OpenAI's chief product officer from June 2024 until October 2025, a stretch that covered some of the company's most aggressive product expansion. Stoke declined to make Weil available for an interview.
The timing invites an obvious question. Sam Altman was reportedly kicking the tires on Stoke last year, weighing a personal investment in a would-be SpaceX competitor, and Weil's board seat is the sort of connective tissue that fuels that kind of speculation. Lapsa waved off the OpenAI chatter as "gossip and rumors," saying Weil is there to focus on Stoke.
The product Weil is now helping oversee is Nova, a rocket designed to be completely reusable and flown repeatedly. No launch provider has yet fielded a fully rapidly reusable vehicle, with SpaceX's Starship coming closest. Reentry heat has historically deterred even the deepest-pocketed space investors from committing to full reuse, and Jeff Bezos' Blue Origin, where Lapsa once worked, has flirted with the approach without prioritizing it.
That calculus has shifted. SpaceX's public debut priced in the assumption that Starship will fly operational missions this year, and the wider market has absorbed the message that rapid reuse is not a fringe idea but the direction of travel. Global launch capacity remains constrained, and the next company to fly a reasonably priced reusable vehicle regularly stands to capture significant demand.
One of the demand vectors Lapsa keeps pointing at is AI infrastructure in orbit. The idea of distributed data centers in space — using constant solar power and skirting political constraints on Earth — has drawn venture interest, but the economics collapse without cheap, repeatable launch. Lapsa said space data centers "really only make sense with full rapid reuse," a framing that positions Stoke's rocket as an AI-adjacent piece of infrastructure rather than a pure launch play.
Weil brings other useful surface area. He served as president of Planet Labs for three years as the satellite imaging company went public in 2021, so the space business is not new to him. He was also one of four tech executives who joined the US Army Reserve to bridge Silicon Valley and the Department of Defense, a background that maps directly onto the military contracts Stoke will need to win to reach scale.
The caveats are hardware caveats. Stoke still has to fly. Full rapid reuse remains unproven at the scale Nova is targeting, and rocket programs routinely slip. Lapsa acknowledged the execution risk plainly, saying the company has "a good chunk of the risk behind us" but more to go, and that Stoke will launch "when it's ready" rather than on a fixed schedule. A director hire does not compress a test campaign.
For the AI market, the more interesting read is directional. The most senior product executive at a frontier lab has taken a board seat at a launch company whose CEO explicitly ties the business case to orbital compute. Whether or not OpenAI ever writes a check into Stoke, the industry is quietly assembling the pieces — chips, models, and now rockets — for a scenario in which AI infrastructure is not confined to terrestrial data centers. The companies laying that groundwork now, at hardware timelines, are the ones that will define what "AI infrastructure" means five years out.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
