This chapter explores the complexities and limitations of large language models like ChatGPT and Copilot, focusing on ethical uses and retrieval augmented generation. It highlights the importance of due diligence in trusting AI-generated content, while sharing personal experiences with these technologies and their potential benefits.
Transcript
chevron_right
Play full episode
chevron_right
Transcript
Episode notes
On this week’s show, Chris, Lisa and Michelle dive deep into Agentic AI, exploring how it differs from traditional automation and current AI tools. Chris explains that while regular agents follow basic rules (like a thermostat) and AI agents respond to commands (like Alexa), Agentic AI systems work autonomously toward outcomes – they’re proactive, learn from mistakes, and can collaborate with other AI agents. The discussion illustrates how these systems can take on complex, ongoing tasks that would be tedious for humans.
The conversation also tackles the significant challenges and risks of Agentic AI implementation. Key concerns include maintaining human accountability when autonomous systems make decisions, the degradation of AI quality due to increased demand and training on AI-generated content, and real-world legal issues like the Workday class action lawsuit involving age discrimination in AI screening.
The AI-powered Podcast Player
Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
