In the full video, I dive deeper into the specific metrics we used, but I also warn about two massive long-term risks that this study hints at...

Watch the full breakdown of the study and what it means for the future of our industry here ➡️ youtu.be/b9EbCb5A408

8/8

The study highlights a critical concept: AI is an amplifier. If you are already applying good engineering discipline, working in small batches, testing, and designing for modularity then AI amplifies your impact.

7/8

Does this mean AI is a magic bullet? Um, no.

6/8

The data showed:
• No big difference in maintenance costs between AI and human-generated code.
• AI users were indeed 30% to 55% faster during the initial phase.
• For experienced devs who used AI habitually, we actually saw a small improvement in maintainability.

5/8

1. Creation: Developers wrote code (some with AI, some without).
2. Maintenance: A different group of developers had to evolve that code, without knowing if it was AI-generated or human-written.

The results were not what many of us expected.

4/8

We decided to stop guessing. I helped run a controlled experiment involving 150 participants.

We split the study into two phases:

3/8

They measure typing speed or feature count short-term metrics that ignore the economic reality of software development. The dominant cost in software is MAINTAINING code. Maintenance represents 50-80% of the total cost of ownership.

2/8

Subscribe & turn on notifications so you don't miss an upload:
youtube.com/@ModernSoftw...