This episode of Techsplainers explores two fundamental capabilities of AI agents: communication and learning. We examine how AI agents exchange information with each other and humans, including agent-to-agent protocols like KQML and FIPA-ACL. We also look at the challenges they face with standardization, ambiguity, latency, and security. The discussion then shifts to how agents learn and improve over time, covering supervised learning with labeled data, unsupervised learning that finds patterns without human oversight, and reinforcement learning through trial and error with rewards. We also explore continuous learning, where agents adapt to new information without forgetting previous knowledge, and how these capabilities combine in multi-agent systems to create collaborative intelligence that can solve complex problems across various industries.

Find more information at https://www.ibm.com/think/podcasts/techsplainers.

Narrated by Selma Pacheco Jimenez