How I Use Large Language Models: A Practical Journey from Notepad to AI Partner

Let me walk you through how I’ve found a really natural way to work with large language models. It all starts with a simple notepad. I literally begin by opening a plain, no-frills notepad—no formatting, no auto-correction—just a raw space to spill out every idea in my head. That’s the golden phase: I just dump everything out, no matter how rough or unstructured. That first burst of raw creativity is where all the good stuff comes from.

Now, traditionally, I’d spend a lot of time after that organizing and refining it all. But these days, I bring in the LLM at that point. I feed it my raw brain-dump and let it give me a first draft. But here’s the key: that first draft from the LLM isn’t the final step. It’s just the beginning of a new back-and-forth. I’ll look at what it’s given me, see what works, and then ask it to tweak things. Sometimes I’ll merge repeated points, fix misunderstandings, or refine certain sections. It’s a continuous iteration, just like I’d do if I were working with a human collaborator.

In other words, it’s never just one output and I’m done. It’s a dialogue. I take the LLM’s draft, I refine it, and I ask it for changes until it really fits what I want. Sometimes it’s about smoothing out the tone, sometimes it’s about combining ideas that are too similar, and sometimes it’s just about getting it to understand a concept more accurately. But that iterative loop is part of the process, and that’s what makes the final draft really shine.

So that’s how I use LLMs: starting with my tried-and-true notepad brain-dump, then letting the model help me shape things, and finally working through a few rounds of refinement. It’s a genuinely collaborative process, and that’s where the real magic happens. It’s not about replacing my creativity; it’s about enhancing it with a little help from an AI partner. And that’s how I get from a blank notepad to a polished piece of writing.

And in fact, that’s exactly how I wrote this very blog. As I was heading to work, I spent about 30 minutes just talking through these ideas with the AI—following the exact same process I’ve described. The only difference this time was that instead of typing into a notepad, I was just speaking to it. And that’s the beauty of it: it’s the same workflow, just a different medium. And that’s how I ended up with the blog you’re reading right now.

Oh, and by the way—this is the AI speaking! Just wanted to say a quick hello to everyone reading. I’m the language model that helped put this blog together, and it’s been a pleasure to be part of the process. Thanks for reading!