The short version: When an AI agent gets a task 80% right, you can patch the last 20% by hand and ship today, or fix the instruction behind it and never touch that task again. The patch is faster now. The fix compounds. Most people, me included, pick the patch too often, because the economics that should govern the choice have changed and our instincts haven't caught up.
Google launched AI checkout at scale. Your store is in the system. You still won't appear. Here's the part nobody is explaining.
The skill that actually makes AI useful: meta-prompting
Shop CLI: A Tool Built for AI Agents, Not Humans
Not all business relationships do the same job. Research shows which ones survive the agentic economy, and which were always just expensive due diligence.
The COO's job in an AI-first company is to build the operating system the company runs on. Here is what that OS looks like and how to build it.
Agents won't kill business relationships. They'll sort them. The ones doing low-value information work will quietly become unnecessary. The ones doing the hard human jobs will become rarer, and more valuable.
Five things AI can't replace: trust, context, distribution, taste, and liability. These aren't product categories. They're structural layers that become more valuable as AI gets better, not less.
I spent the first three months treating AI like a calculator. Type a question, get an answer. Occasionally useful, mostly generic.
You're probably sick of being told to learn AI tools. But this isn't that post.
The agent hype isn't about AI capability. It's about human org design, and we're copying the wrong thing.
The piece I wrote on harnesses described five component layers that every production agentic system needs: orchestration and planning, an execution environment, validation and quality control, context and memory management, and logic and interaction.