The Post-Coding Era: What Happens When AI Writes the System?

Nicholas Moy, former Head of Research at Windsurf (formerly Codeium) now at Google DeepMind, joins High Signal to discuss the rapid evolution of AI-powered software engineering. We discuss Windsurf's journey from early prototypes that struggled with compounding errors to the successful launch of their agentic coding product. Nick explains that building a startup in the current climate requires a strategy of "disrupting yourself" to avoid the innovator’s dilemma; companies must be ready to pivot as soon as a new frontier model makes previously impossible features viable. He argues that traditional technical moats are increasingly fragile, and true defensibility now comes from real-world usage data, brand reputation, and a deep intuition for what users need at the frontier of these capabilities. The conversation also explores the shift from "co-driving" to a truly "agentic" era of development. While early AI tools required developers to monitor every line of code in real-time, the next phase involves assigning projects to agents that work independently and present their results for review later. Nick highlights the engineering challenge of writing software that remains useful even as the underlying models improve every few months, suggesting that builders should focus on "invariants" (the fundamental information needs and output forms of a task) rather than brittle, model-specific harnesses. 00:00 Introduction to Agentic AI in Software Development 01:21 Exploring the Shift from Co-Driving to Agentic AI 02:31 The Journey of Building Windsurf 04:33 Challenges and Successes in AI-Powered Coding 06:07 The Future of AI Agents in Software Engineering 08:34 Building and Maintaining a Competitive Edge 13:39 Product and Engineering Challenges in AI Development 16:28 The Evolution of AI Usage in Software Development 24:23 The Economics of AI and Model Training 32:48 The Future of AI Agents 41:05 Conclusion and Final Thoughts