The 100-Agent Problem Nobody Wants to Talk About Boris Cherny, the creator of Claude Code, said something at a recent Sequoia AI session that stopped me mid-scroll. “100% of my code is written by Claude Code. I run around 100 agents at one time.” He’s an Anthropic engineer reportedly compensated at $750K per year. He…
Most AI agents are blind. They see screenshots. They guess at selectors. They retry five times when a button moves two pixels. That is not a browser. That is a blindfolded person poking at a touchscreen. The dev-browser approach changes the mental model entirely. Instead of giving an AI agent a camera pointed at Chrome,…
Container Pooling in the Responses API Is Not a Plumbing Detail Most engineers I talk to treat cold-start latency as a footnote. Something to optimize later. A known cost of doing business with containerized infrastructure. I’ve been guilty of this too. OpenAI just made that attitude a lot harder to defend. They shipped container pooling…
421 Pages. Free. No Catch. This Google Engineer Just Dropped the Agentic AI Curriculum the Field Has Been Missing. There’s a certain kind of technical resource that only shows up a few times a decade. The kind that was clearly written for internal use, then someone decided the world should have it. This week, a…
The Seams Are Where Your Agent Breaks I’ve shipped enough AI agents to know where the bodies are buried. And they’re almost never where you expect them. Everyone obsesses over the reasoning loop. Prompt engineering, retrieval quality, context window strategy, model selection. All real concerns. All worth the time. But in my experience, the 90%…
Apple Just Weaponized Xcode Against the AI Coding App Market This one deserves more attention than it’s getting. On February 27, Apple pushed Xcode 26.3 to the Mac App Store. Greg Joswiak announced it directly: “Xcode 26.3 with Claude Agent and Codex hits the Mac App Store today! With advanced reasoning capabilities in Xcode, you…
