Ask any tool to write your follow-up email and you’ll have one in four seconds. It’ll be grammatical. It’ll be structured. It might even be good. Then you read it, and you start changing things. The opening is too stiff. It doesn’t mention what you discussed on the call. The ask is buried. The tone is slightly off for this particular client who hates exclamation points. Five minutes later you’ve rewritten most of it. The AI saved you the blank page and almost nothing else. Th

Open most operations job descriptions and you’ll find a list of tasks. Manage the inbox. Schedule meetings. Update the CRM. Draft reports. Coordinate with vendors. That list made sense when each task required a person to physically do it, start to finish. It makes a lot less sense when AI can draft the report, schedule the meeting, and flag the CRM updates that actually need a human decision—leaving the person in that role with a job description that no longer describes their

There’s a particular kind of LinkedIn post making the rounds again. The one where a founder claims they’ve replaced their entire ops team with an AI agent and a few Zaps. Revenue is up, headcount is down, and they’re enjoying their morning espresso while the bots handle everything. What’s missing from that story is what happens at 4pm on Thursday when something breaks. AI is genuinely good at a lot of operations work now. Drafting standard responses. Pulling data. Maintaining