Ashlee Vance launches Core Memory: tech TV show, two books, and the future of independent journalism
Mar 13, 2025 with Ashlee Vance
Key Points
- Ashlee Vance launches Core Memory, an independent media company covering frontier technology, with 15,000 Substack subscribers and a TV show debuting the day after this interview.
- Vance is publishing two books simultaneously through Penguin Random House: one on OpenAI and AI with movie rights already sold, plus a second unannounced title researched for two to three years.
- Vance plans to produce 20 to 30 single-company episodes and two full documentary seasons within the year, targeting a niche audience that wants technical depth rather than mainstream smoothing.
Summary
Ashlee Vance has left Bloomberg to build Core Memory, an independent media company covering frontier technology — space, biotech, brain-computer interfaces, and adjacent fields. The venture is moving fast: 15,000 Substack subscribers already, a TV show launching its first episode the day after this conversation, and a production target of 20 to 30 single-company episodes plus two full documentary seasons of four hour-long episodes each, all within the year.
The TV show is a direct successor to Hello World, the tech travel show Vance made at Bloomberg that earned the outlet its first Emmy nomination. The format is the same — on-site visits, factory floors, founders explaining what they're actually building — but the editorial mandate is deliberately narrower. Vance told his editors to let technical conversations breathe, whether that's a founder nerding out on a rocket engine or explaining stem cell engineering. The explicit target audience is people who want to learn, not a mainstream Bloomberg audience that needed ideas smoothed down.
Distribution and business model
Vance is treating the early release period as an audience-building exercise, letting content out freely before moving toward a subscriber-first window. The plan is to give Substack subscribers early access to videos, release some later on YouTube, and keep the deepest documentary work exclusively on Core Memory. Whether the business runs on subscriptions alone or needs sponsorship is still open — he's explicit that he needs to build the audience before he can answer it.
The scripted work he's developing is the highest-upside scenario financially, though he frames it as "winning the lotto." The baseline goal is sustainability: enough revenue to keep making what he wants to make.
Two books in progress
Vance is simultaneously working on two books. The first, to be published by Penguin Random House, covers OpenAI and AI broadly — he's been reporting it for 18 months and has already sold the movie rights. The second is unannounced but has been researched quietly for two to three years.
On legacy media
Vance is candid that going independent isn't for everyone — he's not certain he can pull it off himself. What he does know is that traditional platforms shaped his voice in ways he didn't fully notice until he left. His Register voice, which he describes as his happiest writing register, came back immediately once he was outside Bloomberg's editorial structure. The comparison he reaches for is his early career writing for The Register, a British tech site that let writers say what they actually thought.
He doesn't rule out a future partnership with a legacy distributor — the Pat McAfee model, where an independent creator eventually distributes through ESPN while keeping creative control — but flags that Bloomberg showed little openness to that kind of arrangement when he tried to have the conversation from inside.
Humanoid robotics
Vance has visited Figure and Xan (likely Agility or a similar company — the transcript is noisy here) and is cautiously more optimistic than he expected to be. Figure's progress over 18 to 24 months exceeded what he'd have predicted. He was initially skeptical of Tesla's Optimus, thinking it was heavy on smoke and mirrors, but acknowledges the progress has been real. The scar tissue comes from Asimo, Baxter, and Boston Dynamics hype cycles that never translated into practical deployment — movement, he says, always proves harder than it looks. He's watching, but not calling a timeline.