Artificial intelligence has become one of the hottest topics in software engineering, but separating genuine opportunities from unrealistic expectations isn't always easy. In this interview from the Software Architecture Gathering 2025, Vaughn Vernon, software ecologist, architect and modeler, reflects on his own journey from early skepticism to practical adoption, explaining how AI can improve developer productivity while emphasizing that craftsmanship, domain knowledge, and critical thinking remain irreplaceable.

Key Takeaways:

1. AI is most valuable when it accelerates software development without replacing human expertise.
2. Large language models can produce convincing but incorrect answers, making human verification essential.
3. Developers should use AI to complement their skills, not as a substitute for craftsmanship.
4. The cost of AI should be evaluated against the productivity gains it can provide.
5. Even when working with AI, professionalism and a human mindset still matter.

Vaughn Vernon offers a refreshingly balanced perspective on AI in software development, avoiding both hype and fear. Instead, he explains how developers can benefit from modern AI tools while continuing to build the knowledge, judgment, and craftsmanship that great software architecture requires. Watch the full interview below to hear his practical insights and real-world experiences.