10+ hours of podcasts, in 5 minutes.
5 quotes from 1 episode on How I Invest, each with a timestamped link to the source.
John Borthwick argues that open-source AI companies succeed by building infrastructure beneath the application layer. Hugging Face became the dominant open-source model marketplace by transitioning from a consumer product into a platform for language models.
Early conversational technology worked best when developers restricted interfaces to a single vertical category.
“Our idea was at Beta Works was that with Poncho that the technology was well suited to a narrow experience, a vertical narrow experience…”
John Borthwick dismisses abstract fears about AI systems taking over the world.
“Am I concerned about some abstract existential risk of AI overtaking the world? I'm not…”
From Why AI Sandbox Escapes Threaten Legacy Software Infrastructure
Hugging Face started as a consumer application before transitioning into a foundational marketplace for language models. John Borthwick describes this shift as moving deeper into the technology stack.
“The pivot that they made to becoming a marketplace for language models, I viewed it as that they were kind of going deeper in the stack and moving from being an app to essentially being a marketplace for models underneath.”
The transformer architecture and the BERT paper allowed developers to build generalized systems. Hugging Face used these research papers to release an open-source language model library.
“When the transformer paper dropped and then when BERT dropped, which was a first instantiation of a transformer-based language model that they could work with, they basically created an open-source version of BERT…”
John Borthwick states that safety audits for frontier models rely on independent testing firms. The internal laboratories building the systems cannot accurately conduct safety tests on their own work.
“First of all this stuff has to be like truly independent of the labs…”
From Why AI Sandbox Escapes Threaten Legacy Software Infrastructure
How we attribute quotes. Every quote was matched against the episode transcript, so the words and the timestamp are real (we trim filler words like "um", nothing else). The name comes from our written summary of the episode, and we use it only when a separate check of the captions finds that person on the episode. YouTube gives us no voice-by-voice transcript, so open the timestamp to hear who is talking. See a wrong name? Tell us and we fix or remove it.
The week's private equity podcasts, boiled down to one short Sunday read, for deal professionals, operating partners, and LPs.
For now, every subscriber gets both newsletters. No spam. Unsubscribe with one click.