Journal

Sometimes I think about how underrated power outlets are.

Hear me out. Whatever appliance, machine, or electronic device you own was probably built by a different company, in a different time, for a different purpose. Whether it’s your lamp or your custom built home lab, the cords still end up on the same plug in the wall.

And nobody rewires their house when they buy a new lamp.

MCP servers are the outlets for AI. They allow any model, or agent, to get plugged into a specific service. It may be QuickBooks, HubSpot, or your internal dashboards. In each instance the wiring behind the wall is different, but the plug is not.

This means that I can connect Claude to my financials and generate the freshest QBR report, while having Astra create a beautiful dashboard to display the results. Two different models, two different jobs, one source of truth.

And when the next best model inevitably comes out next week, I don’t have to rewire anything. I simply unplug one model, and plug the new one in.

If you are interested in learning more about MCP servers, or curious to learn more about AI in general, feel free to reach out to me!

Does anyone else love having a dishwasher?

Instead of having to wait for the water to heat up, soak the dishes, and hand wash with soap, I now load the dishwasher, press start, and it runs the same cycle it ran yesterday.

That's the part most people miss about working with AI.

When someone first begins, every task is hand washing. The whole workflow from scratch, every day. Here's the format I want, here's the tone, here's the thing you got wrong last time, here's the file, here's what to skip. The worst part is the next day you do it again the same way, over and over.

AI Skills are like the appliances you find in your house. You define the process once, and it runs the same way every time. The model still writes the words, but the output is already formed. Same cycle, same expectations, no unnecessary setup every time you want to get some work done.

Each appliance in your house is built with a unique purpose in mind. One might be to clean dishes, or another is to reheat food. Either way, different appliances have different outputs. Skills operate the same way, you build a narrow one that does a specific job well. The power isn't in one machine that does everything. It's in a kitchen full of machines that each do one thing really well.

If you're still hand washing every task, message me. This is a fixable problem, and usually a fast one.

When you think of AI, what is the first thought that pops into your head?

It may be the fear of being left behind, the uncertainty the future holds, or “ugh, another AI slop post”. For me however, mine is different. I want to help teach the basics of AI to those that may not fully get it.

Let's take one of the terms that trips people up the most: API keys.

Everyone knows a garden hose. You walk around the side of the house, twist the spigot a few turns, and water comes out. Then it's yours to use however you want: water the garden, rinse the driveway, or ambush my siblings (not my proudest moment).

An API is the plumbing. It's the line connecting you to something powerful you didn't build yourself. Whether that's a large language model, a payment system, or a financial database. The connection is already there.

An API key is the handle on the spigot. It's what proves the water is yours to turn on. No key, no flow. The hose just sits there.

That's the whole concept. "Application Programming Interface (API)" sounds like it requires a CS degree. It doesn't. If you can turn on a hose, you already understand it.

If you are interested in learning more about API keys, or curious about AI in general, please contact me. I would love to talk!