What I noticed
My mind works faster and better than my typing
It takes me back to school, where essays had to be written according to the “key” the teacher would use. Now it actually doesn’t matter, but this limitation remains deep in my thinking.
The more I speak, the more I realise I’m not thinking faster, but better
That way I can free my mind and simply go with the flow. When I cannot explain something verbally, maybe the idea is not simple enough yet to pass on.
I may not articulate it well, but the main idea is understood better
Simplicity is the key. Good ideas just click. You feel them in your gut. When I have to over-explain any of my ideas, I feel like they didn’t come across at the beginning.
I need to feel that I’m creating something I can call “mine” in the end
This is the most difficult thing to put into simple words, but that’s how I operate. Whenever I pick up a new project, I need clear motivation to pursue it. Even when I’m collecting a lot of data that I will never read, I want to have some common ground.
I don’t want to make anything up or let AI decide. My project has to feel like something I crafted, even though AI helps make it come true.
My best master prompts
My best master prompts that became real projects were not delivered in one go. They did not follow strict rules. I did not try to rerun them every time I added something to the formula. I just went with the flow of the conversation inside one chat. When the context dropped, I switched to a new one.
That is how I create something I like to call The Base. It helps me, in the end, turn a three-hour conversation into a simple statement of a few words. That statement is then described further and divided into chapters and specifications. It works very well, and each time I feel good about the results.
There are times when I throw the whole idea into the trash can
This is perfectly normal. Not all ideas need visualisation or a well-crafted description. Sometimes they just exist in one conversation and are later discarded.
But each of these trash-talks has at least one benefit. I do my own research and explore things I haven’t thought about that deeply before. It’s useful, especially for my further work. I try to gather all those “golden moments” and learn from them, still staying curious about the subject.
What I take from it
Better formatting does not mean a better result.
Rather than focusing on the best prompt that will give me the best result, I like to give the most raw input I can. That way, when AI briefly summarises the content, I can still manage it before it grows too much. For me, that is quite important because later on I can pinpoint what still bothers me and feels off.
To be honest, I like how supportive AI can be. All those messages like “that’s powerful”, “that’s rare”, “not everyone gets there” hype me up a little bit. I still look at this critically. When something is overhyped by AI, I challenge it and explain again until the summary reflects what I mean.
I need to feel that the thing I’m creating still feels mine. This allows me to enter the flow state that gives me much better results than a well-composed message to the bot.