Episode 48: AI interoperability and MCP

In this conversation, Josh and James discuss AI interoperability and the concept of MCP (Model Context Protocol). They explore the initial skepticism surrounding MCP, the process of building custom MCP servers, and the importance of permissions and security in AI integrations. The discussion also touches on the future of AI integration in various applications, particularly in shopping experiences, and the potential for these technologies to revolutionise how we interact with digital tools. MCP is a new way to integrate AI with existing tools. Initial skepticism can lead to valuable discoveries in technology. Building custom MCP servers can enhance personal productivity. Permissions and security are crucial when using AI tools. MCP has the potential to go mainstream with the right ecosystem. AI can act on behalf of users, changing the way we interact with technology. Real-world applications of MCP can simplify everyday tasks. The future of shopping may involve AI-driven experiences. AI tools need to be intuitive for non-developers to adopt them. The evolution of technology often leads to a blend of old and new experiences.