Daniel Kang, professor of computer science at UIUC back on the show. Daniel and Anna have been doing a project together around zkpod.ai. They aim to use zksnarks to prove that some piece of audio is from us,. That is from the original podcast audio. This is similar to his previous work on image provenance, now applied to audio. And so this way you can prove that you produce a specific clip of audio using a real microphone.
In this bonus episode, Anna jumps back on the mic for a quick follow-up to Episode 279: Intro to zkpod.ai. Guest Daniel Kang describes a joint project he is doing with zkpod.ai - the Attested Audio Experiment. The objective is to prove that some piece of audio originates from a "real" (non-ai-generated) recording, even if this audio has been altered through edits or modifications. The goal is to differentiate between original "real" audio and the audio generated by zkpod.ai. In the future some iteration of this architecture could be used by content creators to distinguish deepfake audio from the real thing.
Here is the write up about the Attested Audio Experiment on Daniel Kang's blog.
Here is some of his other work as well:
Follow @zkpodai on twitter for more updates!