
February 18, 2026
Pipes All the Way Down
Greg Isenberg declares the Skill Era of the Internet. The frame is good. The analogy to Stripe-style infrastructure is where it falls apart.
A dedicated publishing area for weekly AI signal: commentary, public opinions, and practical analysis for professionals making decisions.

February 18, 2026
Greg Isenberg declares the Skill Era of the Internet. The frame is good. The analogy to Stripe-style infrastructure is where it falls apart.

February 18, 2026
Claire Vo's same-day heuristic for organisational speed is useful. The FOMO framing and fire-everyone prescription around it is not.

February 18, 2026
A viral post calls out fake AI agent content — while running the same playbook. On the credibility game that every hype cycle produces.

February 18, 2026
OpenForage's AI swarm design document describes an interesting ensemble theory — wrapped around a centralised hedge fund with an API.

February 18, 2026
A crypto trader's viral AI class-war thread cites real data, then builds a sales funnel on top of it. On the gap between evidence and conclusion.

February 17, 2026
The real story in Anthropic's Sonnet 4.6 launch isn't the model — it's what happens when frontier capability migrates down the price curve.

February 17, 2026
Nicolas Bustamante's framework for vertical software disruption is useful — but the founder building the disruptor has skin in the game, and the last 20% of capability is where the liability lives.

February 17, 2026
Will Manidis writes beautifully about parishes, pilgrimages, and the digital economy — but the essay's conclusion about scarcity and meaning flatters a very specific tax bracket.

February 17, 2026
Flynn argues AI agents collapse transaction costs so dramatically that commerce rewires itself. The Coase logic is sound. The infrastructure data tells a different story.

February 16, 2026
Rohit's production agent framework is sound engineering advice. But 40% of agentic projects fail, and the article never asks whether yours should exist at all.

February 16, 2026
Andrew Yang predicts 20-50% of office workers get cut within years. He's good at naming the problem. The solutions need more than a Substack post.

February 16, 2026
Kyzo predicts AI agents will kill Splitwise in six months. The infrastructure insight is real. The timeline is a dinner receipt scaled to civilisational prophecy.
Subscribe to the AI Primer Briefing for weekly commentary on what is happening in AI and what it means for your work.