Colin is joined by Dean Kerr (Lead Developer) and Amy Laws (Developer) to discuss 'The Experiment' – a four‑week study exploring how AI really affects software development. Instead of synthetic benchmarks, the team tackled genuine issues in an open‑source project, alternating between AI‑assisted work and going completely 'cold turkey'.

The contrasts were striking. Amy's AI‑free period exposed how dependent everyday development has become on instant summaries and contextual answers. Meanwhile, Dean found that adopting a simple agentic loop – Analysis → Implementation → Reflection – helped him make better use of AI rather than blindly accept its output.

Together, their experiences reveal a discipline in flux: developers gain speed and support from AI, but also confront questions about craftsmanship, learning, and where the fun in software now sits as the tools reshape both workflow and mindset.

Useful links for this episode

Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity – METR
SWE-bench – SWE-bench Team
Analysis → Implementation → Reflection – a practical technique for issue resolution with agentic AI – Dean Kerr, Scott Logic
Agentic Engineering Patterns – Red/green TDD – Simon Willison's Weblog

Is AI taking the fun out of software development?