Stop Explaining Yourself Every Time

I hope this message finds you well, unlike my AI session last Tuesday, which found me explaining myself from scratch for the fourth time that week.

That is what this piece is about.

Brief First

I use Claude, Gemini, and ChatGPT without a paid subscription on any of them. That is not a flex. That is just where I am in the process. I am in the learning phase, and during the learning phase, the free tier is enough. When I know what I am actually doing, I will pay for what I need. Paying before you understand what you are buying is expensive confusion.

The problem with free tiers is that they do not remember you. Close the tab, lose the context. Every session starts cold. The AI does not know who you are, what you are building, or how you think. You spend the first ten minutes re-explaining yourself and wonder why the output feels generic.

The fix I use is a briefing doc.

I keep mine on Google Drive. Free. I have several, one for each context I work in. Before I open a new session on any platform, I open the relevant doc, copy the contents, and paste it in as my first message. The AI reads it. Context established. We can start.

This is not a complicated system. A Google doc, opened before the session begins. That is the whole thing.

But here is where people get it wrong.

The first version of my briefing doc leaned hard on the fact that I spent eight years as a professional poker player. I included it because I wanted the AI to read me as someone who thinks analytically. What I got instead was an AI that peppered every response with tells, bluffs, and reading the table. It captured the atmosphere and missed the point entirely.

The doc had the right idea and the wrong execution. Poker is where the analytical thinking came from. It is not what I want the AI to reflect back at me. The fix was straightforward once I saw it. Instead of listing poker as background, I wrote what I actually wanted: analytical thinking, empirical reasoning, secondary consideration before conclusion. The source of the trait is mine to know. What the AI needs is the trait itself.

A briefing doc is not a vibe sheet. It is a set of instructions that produce the output you actually want. The more precise the doc, the sharper the session.

The last thing to know: the doc is never finished. Mine has changed several times and will change again. As you learn how the AI reads your instructions, you find the words that work and cut the ones that do not. It is a working document. There is no final version.

Start with one doc. One session. See what comes back. Adjust.

You do not need a subscription to do this. You need a Google account, which you already have, and twenty minutes to write something honest about how you work and what you want.

Do that before you pay for anything.

V> The source of the trait is yours to know. What the AI needs is the trait itself.