Code Is Cheap. Trust and Compute Aren’t.
About this session
How AI-native software development moves the bottleneck from code generation to verification, release engineering, and infrastructure capacity.
AI coding agents are changing software engineering at an extraordinary pace. Generating code is becoming cheaper and faster, and increasingly, AI is also being used to review, test, and fix the code that other AI systems generate.
But this creates two new problems.
First: if AI writes the code and AI reviews the code, who reviews the AI reviewer?
Second: even if we can generate software almost without limit, the infrastructure that builds, tests, deploys, and runs it is still finite.
This shifts the engineering bottleneck downstream.
The challenge is no longer simply how fast can we write code? It becomes: What should we build? What should we trust? What should we release? When should it run? And should it consume scarce infrastructure capacity at all?
Speaker
Key takeaways
- The future SDLC is about governing machine speed. Successful engineering organizations will need to combine AI agents with strong release gates, infrastructure awareness, observability, rollback, and clearly defined human accountability
- Compute is finite even when code generation feels infinite.