Practical AI Exposing the deception of DeepFakes
Jun 25, 2019
Delve into the unsettling world of deepfake technology and its dangerous implications for truth in media. Learn how these manipulated videos can influence public perception and pose risks in political contexts. The discussion highlights the challenges of detecting deepfakes and the need for legislative measures to combat misinformation. Gain insights on safeguarding against these threats while exploring the dual nature of AI technology in both positive and negative applications. It's a vital conversation about a modern menace that everyone should hear.
AI Snips
Chapters
Transcript
Episode notes
Deepfakes and Deep Learning
- Deepfakes leverage deep learning models, primarily GANs, to manipulate or generate media.
- Recent advancements have made creating deepfakes easier, requiring less expertise and resources.
Deepfakes in the Media
- A deepfake video of House Speaker Nancy Pelosi made her appear to have slurred speech, impacting public perception.
- Another deepfake showed Mark Zuckerberg saying things he never actually said.
The Danger of Accessibility
- Deepfakes exploit readily available data from social media and smartphones, targeting anyone.
- This creates opportunities for personal harassment and large-scale societal manipulation.