Well, first of all you will have to figure out what you want to do. You have to have a goal and you have to understand where you are right now and to get to that goal you will need a process. And if you can systemize this process then it’s even better because then you can automate it or delegate it.

So first things first, rely heavily on the goals of the niche or the process you are working on. I spent two years writing this blog related to marketing and stuff, and it was still incomplete in a sense when I went through the boron letters. That book was like two three years ahead of me if I had. kept the same pace, and when I made the complete system, I required like thirteen books and the lessons from it. So you can imagine for the next thirty forty years I would have been a clueless monkey writing on a blog.

So first of all, rely heavily on the goats, rely heavily on the niche books. Avoid popular books, they are slop, they are high time, they don’t get shit done. Look at the practitioners, look at the goats, look at niche things. They have the real value and tactics and mechanics of things.

Once you go through all these books, understand what they are trying to do at what step, and figure out what is going to be the order of steps in the process, and how lessons from these goats can be incorporated into those.

Okay then, once you have the rough skeleton, do a dry run.That will show you the end product and you will realize how far you are from what you want. Now comes the algorithm.

You have the present and you have the ideal that you want, right? So there is going to be a gap between those. So there is present and there is goal, and you are not at that goal right now because of some bottleneck. Maybe you don’t know what next to do. Maybe you know what next to do but you don’t have the equipment for it or whatever. So between you and the goal there are bottlenecks, and each bottleneck you solve. pushes you closer to the goal. So the algorithm is figure out what the goal is, figure out where you are presently standing, what is the bottleneck to the next step. Okay, only the next step, and then when you figure that bottleneck out, think about the best current move. What is the best current move that will either tell you more about the bottleneck or let you move around it or eliminate it? Think in terms of do you want To learn more, do you want to experiment more? Do you want to test things? Blah blah blah, whatever. Figure out the best current move, execute it. After execution, analyze where you stand compared to previous position, and then understand the current bottleneck and go for the best current move again. So it’s like iteration between bottleneck and best current move until you reach your goal.

So this algorithm run will keep guiding you towards your goal. Okay, so when you do the dry run with your skeleton system, you will realize where it is falling short. Okay, so you have to first do the diagnosis of your system where things are lacking and then run the algorithm there.

And after a lot of iteration, your system will become good enough or sub process of it will be good enough. And now you realize that you have made a chain of order, right? Like step one happens, then step two happens, then step three happens. But if step two is really solid now and you’re confident in its output, why always run the step two, right? Just mass produce the output of step two in a reasonable sense. Maintaining its quality so that next time you don’t have to run the step two you can just reach out to the inventory of step two.

And this is how you slowly retire your system because if you are in the content game, there are only so many culturally popular objects that you can use as surrogates, right? For your niche, there might be hundred to five hundred objects. Once you have made the list of those, you don’t need to run that step anymore. It’s redundant. You just have to have that reservoir. And when you keep generating reservoirs for your steps, you unlock the medici effect that may be. The combination that you thought between step two and step four’s output, maybe this five hundred into five hundred can create some new combination you never expected, right? So it’s like once you have the system dialed down, having it produce a lot of things can push your creativity ahead. And this is especially cool for things that are not physical, right? We are I deal with mostly code and information. So for me this. System works perfectly because I don’t have to carry inventory, I don’t have to warehouse it. There is no fucking cost for me. So it instead stabilizes the whole system because who knows that maybe in future there might be prompt drift and what are quality of output I am getting right now, what if I do not get it in future? So it’s better to hold them right now, right?

So until now this thing is set up the goal, then figure out the goals of this, borrow their knowledge, create a rough system, do a dry run, iterate until it is good. Once it is good, create huge reservoirs for each step.

And once you have those reservoirs, just keep running this production cycle and keep making connections and keep sense making of things. That hey, I made this combination but does it make sense?

For example, you are launching a clothing brand. Okay, you can go to the masters of making a jacket. They will tell you all the processes. They will tell you all the components. And then let’s say you are making a T-shirt right now. There is the hem, there are the stitches, there is the front print, there is the back print, and then there is the Bottom hem. I don’t know what it is called. So these are like five six separate objects. You can go and find the culturally iconic objects of it. Like okay, the hem can be ribbed, the hem can be a collar, or it can be chained, whatever. There might be five six culturally iconic objects there. There can be five six variations of stitches that are culturally iconic. So you make an inventory of all those. And now once you have the reservoir of each object. Now you can just go for it, just wing it. You can have like thousands of t-shirt ideas now by just combining things. And yep, that’s how you make a really nice t-shirt. And once you make combinations, you may try to get a sense of it. Like does this actually work? Does my time actually want it?

Once again, rely heavily on the knowledge that already exists. Make your processes good and get output out of them fast because you don’t know how long this process might be good, how fast it might decay because of the reliance on models we have.

Goal → Borrow → Model → Run → Diagnose → Iterate → Stabilize → Stockpile → Compose → Judge → publish.