The Symbiotic SDLC: How AI Broke the Phase Balance
The SDLC was designed for humans. When AI accelerates only some of its phases, the cycle goes out of balance. Production speed rises in development; throughput, stability, and ownership stay the same or fall in QA, deploy, and monitoring. This paper traces the imbalance through delivery data and proposes the symbiotic SDLC as a remedy.
The thesis
AI did not add a phase to the development cycle. It broke the balance between them.
Each phase received an uneven share of acceleration. The numbers from DORA, METR, and Faros AI are consistent: developers feel faster while organizations deliver no faster — sometimes slower. The gap between phases is where entropy accumulates.
This paper is in active development.