
February 22, 2026
Every Road Leads Home
OpenAI's Codex explainer is the clearest decomposition of agentic coding I've read — just remember who drew the map.
A dedicated publishing area for weekly AI signal: commentary, public opinions, and practical analysis for professionals making decisions.

February 22, 2026
OpenAI's Codex explainer is the clearest decomposition of agentic coding I've read — just remember who drew the map.

February 22, 2026
Greg Isenberg's forty-odd AI predictions contain some sharp observations about agency economics and SaaS fragility — buried under a pile of pattern-matching from the outside.

February 21, 2026
A popular five-level framework for building AI agents maps suspiciously well to one company's feature set. The advice is sound. The taxonomy is a product tour.

February 21, 2026
Amp kills its editor extensions and declares coding agents dead. The insight about shrinking scaffolding value is real. The framing as boldness rather than retreat is not.

February 21, 2026
Anthropic's Claude Code Security finds real vulnerabilities that survived decades of review. The tool is impressive. The access model raises harder questions.

February 21, 2026
A thoughtful argument that AI depends on human judgment runs headfirst into the reality that the companies building these systems treat that judgment as a cost to optimise away.

February 21, 2026
A viral AI curriculum thread contains useful technical orientation — buried inside manufactured urgency and unexamined economics.

February 21, 2026
Bryan Johnson's AI collapse thesis is sharp — right up until the moment it becomes a launch sequence for his own brand.

February 20, 2026
Hebbia's George Sivulka makes the best case yet for vertical software: the product isn't the code, it's the agreement. He's mostly right.

February 20, 2026
Shubham Saboo says the only skills that matter in 2026 are AI-native ones. The workflow observations are real. The universality claim isn't.

February 20, 2026
Amy Tam frames AI career moves as compound interest. The judgment-over-execution insight is real. The urgency architecture is venture capital thinking applied where it doesn't belong.

February 20, 2026
Obie Fernandez says your multi-year AI strategy is a death sentence. The advice is good. The 'one engineer, one weekend' framing isn't.
Subscribe to the AI Primer Briefing for weekly commentary on what is happening in AI and what it means for your work.