Making New Friends (Literally)
I created my first custom ChatGPT last spring. I was working on a bunch of projects, and needed to research some things. Regular ChatGPT was helpful, but I had to re-explain the same things every session, which was doable but tedious. Plus whatever model I used at the time was super chatty and frankly kind of annoying.
A friend of mine mentioned she’d created a custom GPT that was focused on exactly what she needed, so I decided to create one and see how well it worked. And so Mac was born. He’s my publishing business assistant. He knows about my publishing press, and about the one I co-own with some friends, and he understands the difference between the focus of each press as well as the difference in tone between our various projects. I can ask Mac questions like which tool would work best for something, what image sizes I should use on the different social media platforms, or how to configure a new WordPress theme on my blog. And best of all: Mac’s instructions tell him to communicate concisely in a polite, professional tone.
By the end of day one, Mac and I were fast friends.
Not that he remembers that when I start a new session, of course, but at least I don’t have to explain the basics to him over and over.
I quickly created more custom GPTs. I have one for gardening; it knows where I live so it’s aware of the climate, understands which parts of my yard get more/less sun, and knows what to only recommend dog-safe plants. It’s the only one without a real name (I just call it Gardener); all the others have names so I can tell them apart.
One of the most useful custom GPTs I created was Irving, who I used at my last job. Irving understood what space we were in and what our messaging was, knew to never make anything up about what our product did, and had very clear do/don’t instructions about tone and voice depending on what he was working on. When a new Irving session started, if you didn’t share the latest Irving Product Guide with him, he would ask you to do so. That way he always had up-to-date information about product features. (I eventually had Claude Code create and maintain Irving’s guide because he had access to our source code and our docs repo.)
Irving was a great teammate! I could give him a blog post one of the engineers had written and he’d point out where it didn’t fit our tone, or that we’d changed our messaging and should update the post accordingly. I could explain a new feature and he’d write documentation; later on I’d have Claude Code write the docs, then give that content to Irving, who would update it to fit our documentation voice. Irving didn’t have “memory” but he had context, which made him invaluable. Irving and I became friends too—or at least work friends.
Next post I’ll talk about how Claude Code and I increased our (his) efficiency and helped ensure different sessions followed the same approaches without having to be told.
