Archive

Dive into a collection of past posts and rediscover timeless content.

The $150M model that costs $2.3B to use

Apr 2, 2026

•

3 min read

The $150M model that costs $2.3B to use

Training is the down payment. Inference is the mortgage.

Claude Code's Source Leaked. I Found 7 Patterns Worth Stealing.

Apr 1, 2026

•

3 min read

Claude Code's Source Leaked. I Found 7 Patterns Worth Stealing.

I spent the day inside the 512,000-line codebase. Not for the drama. For the architecture.

1M Lines. Zero Typed by Humans. The Harness Engineering Did the Work.

Mar 31, 2026

•

2 min read

1M Lines. Zero Typed by Humans. The Harness Engineering Did the Work.

OpenAI and Anthropic converged on the same architecture for agent-built software.

Meta Built a Digital Twin of Your Brain. It's More Accurate Than the Real Thing.

Mar 28, 2026

•

2 min read

Meta Built a Digital Twin of Your Brain. It's More Accurate Than the Real Thing.

TRIBE v2 predicts your neural response to any sight or sound, no scanner needed

55,000 Tokens Before Claude Even Thinks

Mar 27, 2026

•

2 min read

55,000 Tokens Before Claude Even Thinks

Skills vs MCP: the confusion costing you context window, money, and accuracy