Blog & Experiments
Notes on things I've tried, built, or broken. Less formal than articles — more like a lab notebook.
-
May 2026
Using a spec as a benchmark: What 90 model runs taught me about agents, prompts, and shortcuts
I ran the same OpenSpec through 90 runs across models, agents, and configurations. Some results were expected. Others were not.
-
May 2026
AI and the changing structure of work
A conversation with Deepak Tripathi, who leads a Global Capability Center, challenged several of my assumptions about how organizations develop capability. Here is what the research says about where AI is actually putting the pressure.
-
May 2026
If AI is replacing engineers, why are companies still hiring them?
The ATM was supposed to kill bank tellers. It didn't. Understanding why tells you more about AI and software engineering than most takes you'll read on LinkedIn.
-
May 2026
The creative tools question
Anthropic joined the Blender Development Fund, the community pushed back, and BCON made it clear that AI in creative tools is no longer a hypothetical debate.
-
May 2026
Adding MCP is fast. Securing it properly takes intentional design.
90 minutes to add MCP server support to a retrospective board. A new prompt injection attack vector that didn't exist before. MCP changes the threat model, not just the feature list.
-
Apr 2026
AI is not going to make security obsolete. It is going to make it more important.
Anthropic's recent security incident is a reminder that the threat landscape doesn't care how advanced your stack is. The fundamentals still matter, and AI makes them matter more.
-
Apr 2026
I built a retrospective board in under an hour
Testing spec-driven development with modern AI tools, what it revealed about the future of software and what actually matters for developers.