Why AI Won’t Break Brooks’s Law
Fifty years ago, Fred Brooks wrote about the challenges of building large, complex software projects; his most famous assertion: “Adding engineers to a late software project makes it later.”
The digital journal of James Rowe. Something between WordPress, LiveJournal and gists.
Fifty years ago, Fred Brooks wrote about the challenges of building large, complex software projects; his most famous assertion: “Adding engineers to a late software project makes it later.”
Twenty-five years ago, Joel Spolsky wrote that the single worst strategic mistake any software company can make is rewriting the product from scratch. Now, with AI, the prevailing opinion is that companies will clone SaaS products internally, displacing the market.
The two-pizza team isn’t dead—it just needs different toppings. AI breaks previous staffing assumptions by collapsing the time it takes to write code. Engineers equipped with AI can transition coding tasks from ready for dev to ready for test in hours, not days. So how best to leverage this new capacity?
Frontier conversational large language models (LLMs) are the closest implementation of the fictional Star Trek “Computer” I’ve yet to experience. I compare my journey with AI to the evolution of CPUs to GPUs.
This is Part 5 of 5 of my 2025 AI Wrapped series. This post covers key lessons for myself and how I think with AI.
This is Part 4 of 5 of my 2025 AI Wrapped series. This post covers all the ways that AI fails me daily—and why despite the shortcomings I’m still all-in on AI.
This is Part 3 of 5 of my 2025 AI Wrapped series. This post covers what I’ve shipped with 100% AI-generated code. When I first started reflecting on building with AI, this was the first draft. Writing it became the genesis for all the other posts in this series.
This is Part 2 of 5 of my 2025 AI Wrapped series. This post covers my actual development environment setup—the specific tools and workflows I use daily to ship production software with nearly 100% AI-generated code.
This is Part 1 of 5 of my 2025 AI Wrapped series. This post covers my evolution from “CHOP” programming to integrated CLI/Cursor workflows, why I now build prototypes instead of PowerPoints, and how AI makes it possible for engineering managers to stay technical without impeding their teams.
You’ve either asked this question or someone has asked you: “Do you know a good __?” A project needs to be done, and you’re terrified of hiring the wrong person. The contractor who leaves your house half-finished for months. Or, worse, someone who builds a façade of completion only for it to fail inspection or fall apart under real-life use.