We Asked AI to Caption Our Shows - Here's What Happened
In 2020, we asked, “Could cloud speech-to-text APIs caption our shows?” The result was “sort of”- a mix of typos, dodgy accent support, misinterpreted place names, and potential serious liability meant we couldn't comply with required broadcast standards. Fast forward to now when we put the latest AI models to the test and found that thanks to context inference (or by faking it convincingly) things are looking up. Jeremy will share hard lessons, gotchas, and a peek into how these models pick their words that can help get to production-ready.
Adam Brown
Co-Founder & CTO
Mux
Adam Brown co-founded Mux in 2015 and leads technology and architecture for the developer-first video infrastructure platform. With deep roots in video technology, Adam has built high-performance encoding systems, low‑latency live streaming pipelines, and scalable cloud video infrastructure, including during his time at Zencoder and Brightcove, with additional experience in VR rendering at Otoy.
Known for merging engineering rigor with developer empathy, he’s focused on enabling seamless, scalable video delivery and real-time analytics through API-first products like Mux Video and Mux Data.