The Peril and Promise of Synthetic Media | Henry Sabin | TEDxHult Boston

Risks & Challenges - AI technology enables elaborate digital scams, including cloning voices and likenesses to steal tens of millions of dollars [3][4] - Synthetic media can be used for malicious purposes such as creating non-consensual adult content, spreading propaganda, and mass misinformation [8] - The authenticity of online content is increasingly questionable, requiring users to question whether videos are real or synthetic [7][20] Technological Solutions & Countermeasures - Organizations like C2PA are working to standardize content provenance to distinguish between synthetic and authentic content [9] - Blockchain technology can be used to verify the authenticity of videos at scale [9] - Online leaderboards exist for deepfake detection software, fostering competition to create more accurate detection models [10][11] Potential Benefits & Opportunities - AI deepfake technology can translate lectures and videos into different languages, unifying the internet and making content accessible to a wider audience, as over 80% of people don't speak English [12][13][14] - AI video can personalize content to individual viewers, tailoring the level of understanding and removing redundant sections [16] - Personalized AI videos can improve business outcomes, such as increasing patient appointment attendance by five times compared to traditional SMS and email reminders [17] Open-Source Development - AI models are often developed in an open-source manner, allowing anyone to access and build useful applications, preventing monopolization by large institutions [18][19]