A comprehensive 2-hour course on AI agents covering the core agent loop (observe, think, act), setup of three major agentic coding platforms (Codex, Claude Code, and Google's Gemini-based tool), and advanced prompting techniques. Topics include self-modifying agent instructions via persistent markdown files, multi-agent MCP orchestration where different models handle specialized tasks, video-to-action pipelines using Gemini's multimodal capabilities, agent skills for standardized workflows, and stochastic multi-agent consensus. The course targets non-programmers and business users, emphasizing parallelization as the key strength of AI agents and practical patterns for reducing errors over time.
•2h 13m watch time
1 Comment
Sort: