AI Startups

Building Muse, Working with Mark Zuckerberg & The Future of AI | Alexandr Wang, Meta

Oct 11, 2026 · 1:12:17

David Senra interviews Alexandr Wang, Meta's chief AI officer and co-founder of Scale AI, about building Muse, which Wang calls the fastest-growing consumer AI app of all time with 'millions' of users. Muse grew from Nat Friedman's OpenClaw experiments into a one-week prototype demoed to Meta's board in February, then a seven-month build by a team of under 200 to make a reliable personal agent. Wang argues Meta bet on personal AI over coding agents because its heritage and 3.5 billion daily users make consumer AI its edge, and he describes marketing Muse with memes to make Meta cool again. He recounts Zuckerberg's call after the Llama 4 setback, leaving Scale after nine years as Meta took 49%, and rebuilding Meta's AI lab as a flat, anti-bureaucratic team with over 200 direct reports.

How Mark Zuckerberg, Palmer Luckey, Josh Kushner & Scott Wu Think | Jeremy Stern

Oct 4, 2026 · 1:08:11

Colossus editor-in-chief Jeremy Stern joins David Senra to unpack his profiles of Mark Zuckerberg (Meta), Palmer Luckey (Anduril), Scott Wu (Cognition) and Josh Kushner (Thrive Capital), arguing Zuckerberg is more competitive than anyone imagines and will never walk away. An Augustus devotee since 15 who refused Yahoo's $1 billion at 21, Zuckerberg has, Stern argues, the distribution and owned infrastructure to win the AI race. The Luckey profile began with a question — was defense tech real or bullshit after October 7th? — and became 15,000 words on a man who backed Trump when it cost him Facebook. Wu stands out as an authentic math genius who won't sell. Kushner's piece grew from his family's Holocaust story and ends with Stern's Russian professor, who taught him people owe their gifts.

Building One of AI’s Fastest-Growing Companies | Mati Staniszewski, ElevenLabs

Sep 9, 2026 · 1:11:08

ElevenLabs co-founder Mati Staniszewski tells David Senra that voice will become one of the primary ways people interact with AI, and that the company wins by combining frontier audio research with product deployment. He traces the origin to Poland's one-voice movie dubbing he re-experienced in 2021, and to creator feedback that pivoted the dubbing idea into speech generation. He details a Palantir-inspired model: sub-10-person teams, no titles, a five-layer cap, and forward deployed engineers who bring customer learnings back into the product. He cites Deutsche Telekom, Revolut and Klarna, the 20,000-voice marketplace, and voice restoration for over 10,000 people. After three or four acquisition offers, the last in June last year, he and co-founder Piotr Dąbkowski are going all in.