Brennan Builds Things

Recent Posts

Nov 2026
The Trolley Tests Are Out: What 22 AIs Chose, 761,940 Times
How I Built Anatomy of an Agent Harness
Sep 2026
AI Is Changing Your Store
Aug 2026
Anatomy of an Agent Harness
May 2026
I Made This Deck With Codex
Apr 2026
The New Software Factory
Anthropic's Mythos Is On Trend, and That's the Point
The Exact `deslop` Skill I Run Before Commit
The Messy Part of Agent Takeoff Is Starting
Mar 2026
Where the VCs Think All This Is Headed
I Built a Visual Autoresearch Battleground
Feb 2026
Agent Defaults Are the New Market Share
My 7-Year-Old's v0 Usage, Measured
Jan 2026
RAG Ingestion, Rebuilt (and Simpler Than It Sounds)
Dec 2025
Round 2: 17 models vs. my homepage agent
Picking Your Slope on the AI S-Curve
+15 Field Notes
Nov 2025
Product Hunt Suddenly Feels Like a Firehose
+18 Field Notes
Sep 2025
GEPA: From Evals to Shipping Prompts
Coding in 2025 (part ii)
+5 Field Notes
Aug 2025
Prompting Your Prompt's Prompt
Is Turborepo Is the New Collaboration Surface?
The Disposable Intelligence Era
Jul 2025
Calvin's OpenAI Reflections
How I Document Agent Instructions
Jun 2025
Coding in 2025
You. Are. Not. Using. Enough. Compute.
How I Made My Blog Remember You
May 2025
Testing 4o images from the API
Apr 2025
OpenAI vs. Anthropic vs. Google AI
Seeing UV: Detecting Intelligence Beyond Ours
o3 & o4-mini: OpenAI's latest models
GPT-4.1: A true x.1 Release
If AI-2027 Is Right: What Should We Do?
How I Built a Fully-Automated Agent to Promote My Content on 𝕏 and TikTok
Mar 2025
GPT-4o finally gets it's multi-modal moment
A New “Time Horizon” Metric and Doubling Trends
Blogging in 2025
Does Intelligence Make AI Lean Left?
GPT-4.5 is actually a big deal
Feb 2025
Emergent Misalignment: The Hidden Dangers of Narrow Fine-Tuning
Blueprints for the Next-Gen Code Agents
Notes from Deep Dive into LLMs
o1 to o3 - A Bitter Lesson in Coding
What does Claude's usage tell us?
Linear intelligence, superlinear impact?
User’s Attention is All You Need
Jan 2025
Building The Brain