More or Less · Topics · Episode #163
Why OpenAI Will Never Ship a Hardware Device (and the End of the Free-Talent Era)
Clip from the full recording, 47:45–49:37 · download mp4 · watch at this point on YouTube
Sam argues OpenAI's incentives as a soon-to-be-public "narrative capitalist" mean it can never risk shipping a real hardware product, and that the old lab strategy of buying loyalty with unlimited resources dies the moment real revenue numbers show up.
Riffing on reports of leaked Apple hardware and speculation about an OpenAI device, Brit doubts one will ever actually ship. Sam agrees, but for a sharper reason: OpenAI is a “great narrative capitalist” playing “buy the narrative, sell the news,” and a hardware launch is nothing but downside risk once you’re a public company – a bad device tanks the stock in a way a rocket exploding never would for someone like Elon, who’s insulated from public-market judgment on individual failures.
That leads Sam to a broader thesis on why the AI labs’ old talent strategy is ending: for years, labs won researchers by giving them unlimited resources and their own fiefdoms to launch pet projects, essentially paying for talent with narrative rather than discipline. That worked great until the numbers show up – once a company has real, trackable revenue, undisciplined narrative-driven launches become existential risks rather than talent-retention perks. Dave sums up the shift bluntly: “there’s nothing like data to screw up a good story,” and the group jokes that Sam is uniquely under-merchandised as a “meme lord” without narrative-capital t-shirts, spinning off a mock pitch for a venture fund literally named Narrative Capital.
Key points
- Brit doesn't believe OpenAI will ever ship a hardware device, given the leaked Apple device rumors circulating at the same time.
- Sam: any OpenAI hardware launch is "classic buy the narrative, sell the news" -- and pure downside once the company is public, unlike Elon's rocket-failure risk tolerance.
- Sam: the old AI-lab strategy of luring talent with unlimited resources and personal fiefdoms only worked because there was no real revenue yet to measure them against.
- Sam: once a company has numbers, pure narrative-driven talent retention games get much harder to justify internally.
- Dave: "there's nothing like data to screw up a good story" -- summing up why revenue growth is dangerous for narrative-driven companies.
- Running joke: Sam is a self-described "meme lord" with no narrative-capital merchandise, spinning off a mock pitch for a VC fund called Narrative Capital.
Where they landed
Quotes
“It's a classic buy the narrative, sell the news, right? Which again makes these things hard to be public... it will only hurt their stock price.”— Sam
“That worked great until the numbers show up, right? And the problem is now that there are numbers.”— Sam
“There's nothing like data to screw up a good story.”— Dave
Suggested tweets
“It's a classic buy the narrative, sell the news.” Sam Lessin on why OpenAI will probably never ship real hardware once it's public.
Tweet this →Dave Morin: “There's nothing like data to screw up a good story.” The AI labs' free-talent era is ending the moment real revenue shows up.
Tweet this →Someone should really launch a venture fund called Narrative Capital. Sam Lessin is available to design the t-shirts.
Tweet this →All topics · Full episode #163 + transcript · Subscribe on YouTube · Spotify