From Particle Physics to Audio AI with Scott Stephenson - #19

Dark Matter

Scott Stephenson, co-founder and CEO of Deepgram, shares his journey from particle physics to audio AI. His work in particle physics involved searching for dark matter in a lab two miles underground in China, utilizing analog detectors to identify particles. These techniques laid the groundwork for his transition to audio AI, as the methods used in physics were surprisingly applicable to audio processing.

Physics is very, at least particle physics at the very hairy edge of research is still analog, and you still look for particles using analog detectors.

--- Scott Stephenson

Scott's experience in reconstructing 3D events in particle physics provided valuable insights for developing AI-based audio indexing and searching platforms 1 2.

   

Automating AI

The transition from manual data processing in particle physics to automated systems was a significant leap for Scott and his team. They employed machine learning to automate the interpretation of data from photomultiplier tubes, which resemble audio waveforms, to detect particles like dark matter. This automation reduced the need for manual labor in data analysis, a task traditionally performed by scientists.

A lot of this hard, manual labor of figuring out these cuts is done by machine learning now.

--- Scott Stephenson

Deepgram applies similar automation techniques to audio, enabling companies to efficiently analyze customer interactions and improve quality assurance processes 3 4.