Love this, Doug. After-hours is an even better fit for plumbers and electricians here. For them the 10pm burst-pipe call is the job, and whoever answers first gets it. Let's make this the one product we test.

The thread I see running through it is that AI is starting to behave like any other utility. The chips are now political, the models keep improving every month, and Apple charging $50 for usage packs shows the free trial era is ending. For small business owners I think that changes the question. It's less "should I try AI?" and more "which few tools do I lean on, and what are they worth to me each month?" Budget for them the way you would for electricity or a phone line. Then re-test your go-to assistant every few months, because the one you picked in January might not be the best one now. The jobs section is the part I'd underline. Nobody is hiring AI geniuses. They're hiring people who can make the tools behave day to day: a chatbot that answers properly, prompts the staff actually use, drafts turned into posts that sound like the business. That's a very reachable skill for anyone in this group.

Love this, Doug. Keeping the restaurant's own look is the big one. We've been tearing down sites here too and building our own, so I'm going to run your 13-point phone check on a few Gqeberha places

if you want a tip put 50 bucks on #6 in the fifth race😁

Doug — this lines up with what I keep seeing. A lot of people are uneasy about the future right now, and headlines like these make that easy. What I take from your rundown is more practical than scary. The money going into OpenAI and Anthropic means these tools are not a fad, so the skills you build now will still matter. And the jobs showing up are simple, real ones: write a prompt that actually works, set up a basic customer-service helper, or get a newsletter and follow-up running without doing it by hand every time. That is the automation work I care about. It sits beside the job people already do. It does not ask them to wait for a perfect platform, and it does not mean building a custom AI operating system that dies when the next model drops. The Meta “describe what you want and get a little app” direction is interesting. I still think the durable layer is the workflow itself — the automation companies that already run those pipes, like n8n, Zapier, Make and others like them. You can swap the brain later. You should not have to rebuild the plumbing. Your note on the open hacking tools is the other half of it. Do not paste customer info, passwords, or private plant data into a tool unless you know where it is going. Appreciate the Saturday rundown. Johnny