AI has dramatically reduced the cost of writing code, but it has not made software delivery easier. Engineering leaders and architects are now facing a new reality: teams can generate significantly more code than before, yet system quality, coherence, and shared understanding are increasingly difficult to maintain.

In this session, I will discuss how AI is being used in software development inside Microsoft engineering teams for code generation, refactoring, and exploration. While these capabilities unlock productivity, they also shift the primary bottleneck from writing code to owning and governing systems. This creates new challenges: fragmented implementations, inconsistent design patterns, reduced reuse, and a growing gap between output and quality.

This session is grounded in real-world experiences and ongoing internal discussions. I will not present AI adoption as a solved problem. Instead, I will share what is working, where we are struggling, and how teams are adapting their engineering practices in response.

I will address critical questions for modern engineering organizations:
- Is traditional planning still relevant when implementation becomes cheap?
- How do we ensure quality when code volume increases dramatically?
- How do teams maintain shared learning and architectural alignment?

The Age of AI is an evolution, and a major transformation is underway across large organizations. Don’t expect anyone to have all the answers yet; I certainly won’t. What I can offer is an honest view of what we’re learning as we go.

This talk is designed for engineering leaders, software architects, and experienced engineers navigating the next phase of software development.