Jason Howell and Jeff Jarvis discuss the latest AI news, including OpenAI's new Sora video generator, legal issues for an Air Canada chatbot, missteps in publishing AI art in a medical journal, and Jason gives a hands-on demo of Perplexity Pro including a solid example of why it's necessary to scrutinize the output of LLMs.

• Release of OpenAI's Sora for video generation
• Debate over whether advanced AI capabilities indicate nearing AGI
• Meta's new AI model V-JEPA learns from video like LLMs learn from text
• Effect of Sora release on Adobe stock valuation
• Air Canada chatbot provides misleading refund info, company gets sued
• Scientific journal publishes AI-generated rat diagram from Midjourney
• British painter Harold Cohen spent 40+ years refining an image-generating robot

PERPLEXITY PRO DEMONSTRATION

• Integrates multiple models like GPT-3.5 and Stable Diffusion
• Capability to search across internet sources with up-to-date results
• The need to verify its output: Got its own capabilities wrong by claiming it uses model "Sora"

Learn more about your ad choices. Visit megaphone.fm/adchoices