Explore the potential dangers of GPT-3, the world's largest and most powerful natural language model, in this 38-minute Black Hat conference talk. Delve into how this advanced AI tool, developed by OpenAI, can be maliciously used for large-scale disinformation operations. Learn about the capabilities of GPT-3 to generate realistic text across various genres, including news articles indistinguishable from human-written content. Examine the implications of this technology falling into the wrong hands and its potential to disrupt societal fabric. Gain insights from speaker Andrew Lohn on the intersection of artificial intelligence and information warfare, and understand the challenges posed by AI-powered disinformation campaigns.