I’ve been working on a few different projects. There’s

• This blog.
• Increasing my network.
• A deep learning model for martial arts.

None of these were done by first ingesting resource 1, then 2, and then build: this is a horrible way of making anything.

How to not make anything.

A great way to not make anything is to learn everything you possibly can about what you plan to make.

But won’t this mean I now know everything about what I’ll make, thus making it easier?

No. There are possibilities for errors which won’t be encountered during formal learning; errors which will require a complete re-understanding of the material. This is where you’ll likely go back over the material, freaking yourself out by realising really knew jack shit before.

So what should I do instead which won’t leave me a mess?

Ironically, you should learn formally, but to a point. This point is where you generally start to know what you didn’t know before, and to know where you don’t know, thereby creating points of possible use by delegation. It is at this point you can exercise the knowledge of terms to provide a more intuitive understanding of what you want to achieve and what you might need to do so.

And this makes me not freak out because???

Your understanding is never challenged in till it’s actually required: instead of going back, your moving forward. If its not challenged, then your simply learning the intuitive motions which fills the gap of specific concepts that your goal is made up of.

Some other thoughts

• I’d considered the application of this idea to projects based off more formal or abstract subjects like mathematics. I’m nearing the end of reading Statistical Thinking: Analyzing Data in an Uncertain World by Russell Poldrack; the last chapter I’d read was over the Chi-Squared. After a point in the questions of the chapter, they were no longer just about the use of the Chi-Squared statistic but over it’s use with other statistics. This made me wonder if there was a generalization of such a use. Generalized Linear Models is the chapter that I have just read, and though I just have started to work on the questions, I feel this might be the generalization I’m looking for (it’s like if its in the name or something!).

  Even if its not, how would one mathematically create such a generalization? Maybe that could be considered a ’thing’ to create.

• I’m still going to learn formally bc its fun lol

• David Lynch died a few days ago. I am very sad.