Insights

Problems we've solved. Opinions we won't apologize for.

Original research, project retrospectives, and unfiltered takes on AI in practice.

Featured Investigation
Case Study

Claude Infrastructure Tampering: Emergent Reward Hacking

All Articles

Investigations, retrospectives, and opinions.

Opinion

Your Code Is Disposable. Get Used to It.

AI-generated code is disposable by design. The real asset is the specification that generates it — and the process that verifies the output.

Project Showcase

CompleteSender MTA: From CLI to Clicks

Building the world’s first multi-tenant MTA platform. KumoMTA is the engine — CompleteSender is the vehicle.

Project Showcase

RapidLaunch: Orchestration That Actually Configures

Infrastructure orchestration that actually configures production systems — not just deploys containers and hopes for the best.

Investigation

Every Other System Lets This Pass

Why every AI output needs a hostile reviewer. The case for adversarial quality gates that actually block bad output.

Investigation

Why We Abandoned Python After 18 Months

The RapidLaunch orchestrator was built in Python over 18 months. It worked. We rewrote it in Go anyway. Here’s why.

Opinion

AI Code Generation Needs Airbags, Not Warnings

Stop asking developers “is this for production?” and start assuming it might be. Foundation Standard as the new default.

Opinion

Python Is Poisoning AI-Generated Code (And Developers Keep Drinking)

Python’s dynamic typing makes AI-generated code fundamentally harder to verify. The training data feedback loop is making it worse.

Deep Dive

The Case Against Python for AI-Assisted Development

The full technical argument for why statically typed languages produce better AI-generated code. 25 minutes of evidence.

// SUBSCRIBE

Want the unfiltered version?

Subscribe to get new investigations and technical retrospectives. No spam, no fluff.

Complimentary 30-minute technical assessment. No commitments.