Deepfake Detection: Can We Stay Ahead of Scammers? | UC Berkeley PHD Sarah Barrington SVIC Podcast #51

Gain access to our bonus episodes and our private discord community by becoming a YouTube Channel member or going to patreon.com/svic In this episode, we interview Sarah Barrington, a PhD student from UC Berkeley, working with Professor Hany Farid on Digital Forensics, Computer Vision, and AI. A Cambridge graduate with a background in engineering and data science at McLaren, Sarah co-founded two startups, including coyote.ai, which was acquired in 2019. She’s a US-UK Fulbright award recipient and has earned multiple honors for academic excellence. Join our free weekly newsletter so you don't miss all of the tech news: https://svic.substack.com/ Startups: Have questions about being acquired by a large tech company? Jordan has worked on 150 acquisitions during his time at Google and Slack. Reach out to Jordan at https://rebrand.ly/talktojordan Follow us: Threads: www.threads.net/@svicpodcast Twitter: https://x.com/svicpodcast LinkedIn:   / svicpodcast   Tiktok:   / svicpodcast   Instagram:   / svicpodcast   Spotify: https://podcasters.spotify.com/pod/sh... Show hosts Joe (Eng VP) and Jordan (M&A Deal Lead) worked at various companies from Google, Apple, Facebook, Microsoft, Salesforce, Slack, Carta, Splunk, Wealthfront, Adobe, and more. Chapters 00:00 - Introduction and Welcome 00:17 - Podcast Overview by Jordan Thibodeau 00:37 - Meeting Sarah and her Background at McLaren 01:42 - Sarah's Career Pivot to Deepfakes Research at Berkeley 02:12 - Academic Journey: Cambridge, McLaren, and Fulbright Scholarship 03:44 - Early Career: Applying Formula 1 Data Science to Other Industries 05:10 - Transitioning to PhD Research on Machine Learning and Deepfakes 07:27 - Understanding Deepfakes: Problem Scope and Societal Impacts 08:53 - Challenges of Misinformation and Child Protection in Digital Spaces 09:45 - Overview of Sarah's Research Paper on Audio Deepfakes 12:13 - Detecting Voice Cloning: Insights and Approaches 15:50 - Discussion on AI Ethics and Implications of Voice Cloning 16:42 - White House Visit: Policy on AI in Elections and Robocalls 20:24 - Financial Implications of Deepfakes: Scams and Fraud Cases 23:33 - How to Protect Yourself from Deepfake-Driven Scams 27:35 - Latest Research: Human Accuracy in Detecting Deepfakes 30:36 - Closing Thoughts and Upcoming OpenAI Events Single and Multi-Speaker Cloned Voice Detection: From Perceptual to Learned Features: https://arxiv.org/pdf/2307.07683#ai #openai