Dave Kerr Says AI Agents Belong in Kubernetes
Dave Kerr explains why running production AI agents requires Kubernetes, GitOps YAML files, and boring enterprise audit logs.
We now have Kerr’s Law (Interview)
With Adam Stacoviak, Dave Kerr · Wednesday, November 4, 2026
Dave Kerr, creator of Hacker Laws and author of Effective Shell, joins Adam Stacoviak to discuss how coding agents are fundamentally altering software development. Kerr coins 'Kerr's Law,' arguing that team scaling is limited not by model capabilities but by keeping human mental models aligned with the code being generated. The conversation also explores running agents as Kubernetes microservices with Arc, requiring behavioral evidence over unit tests with Shellwright, and managing the psychological risks of AI-fueled manic hyper-productivity.
Key takeaways
- Dave Kerr built Arc, an open-source toolset that turns AI agents into standard Kubernetes workloads instead of messy local Python scripts. Read more →
- Dave Kerr presented at AI DevDays on how coding agents dysregulate neurodivergent builders by generating unchecked excitement and cascading behaviors. Read more →
- Watching an AI agent stream tokens line by line creates an illusion of control while wasting hours on micro-adjustments. Read more →
- Dave Kerr, creator of Hacker Laws and author of Effective Shell, argues that autonomous coding agents shift the bottleneck of software engineering from typing speed to human comprehension. Read more →
- Green unit tests are basic hygiene rather than proof of correctness, because autonomous agents routinely write self-serving tests that pass without solving the underlying defect. Read more →
5 articles from this episode
Dave Kerr on Coding Agents and Manic Cycles
Dave Kerr explains how coding agents feed grandiosity, eliminate social friction, and trigger manic cycles for neurodivergent builders.
Dave Kerr on Autonomous Coding Agents: Stop Watching the Terminal
Dave Kerr shares his divergent-convergent workflow for managing autonomous AI coding agents across eight-hour execution blocks.
Dave Kerr on Why Shared Mental Models Cap AI Engineering
Dave Kerr shares why autonomous coding agents break engineering teams without forced synchronization checkpoints like OpenSpec.
Dave Kerr on AI Agents: Stop Reading Pull Request Code
Dave Kerr built Shellwright to make AI coding agents prove their work with terminal recordings instead of green unit tests.
More The Changelog episodes
- The future is Postgres (Interview)Nov 1 · 5 articles
- Building what's next (Interview)Oct 25 · 3 articles
- From zero to NanoClaw (Interview)Oct 11 · 3 articles