Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn to transform AI-assisted development from unpredictable "vibe coding" into a rigorous architectural approach through this 21-minute conference talk by Ken Wheeler, Ex-Director of Open Source at Formidable. Discover how to leverage generative AI while maintaining control over code quality by combining deep domain expertise with agentic workflows to build verifiable systems. Explore the fundamental differences between noise-generating AI usage and productive "silent shipping," understanding why effective AI prompting requires substantial domain knowledge rather than generic requests. Master three essential rules for escaping "LLM astrology" and producing quality code through proper context and domain expertise. Examine the shift from simple token prediction to deterministic tool usage in agentic workflows, breaking the myth of requiring 4-hour focus blocks for effective development. Follow three detailed case studies including building custom applications in fragmented time, tactical mapping for real-world survival scenarios, and architecting a complete IoT irrigation system. Understand the "trust but verify" protocol for handling hardware hallucinations and learn to implement a practical technology stack using SSH, Tailscale, and Claude Code CLI. Gain insights into real-time system visualization using Grafana on Raspberry Pi, cloud environments, VR coding setups, and the concept of "adjacent coding" for maintaining development momentum across interrupted work sessions.
