Strategy to Consume to Produce with minimal effort
Published on
Getting Things into my Head
I have the goal to Produce, don’t consume and Consume to Produce. But I also have work to do. So how to produce stuff fast? This note explores a technique, which worked quite well recently.

I start by learning interesting stuff. I can do whatever I like Divergent Thinking and follow stuff which is interesting. I am building a Framework in my head and gain a general understanding. But I don’t have to - I can do whatever I want. I just collect stuff.
Getting things out of my head
Then, at some later time, when some threshold is met I dump out my brain. A mind-map without all the rules I read online works well. Externalizing and seeing the result of my messing head usually develops a bunch of new ideas and connections. An Infinity Canvas to Think works well for this, as the mind-map can grow in any direction and usually becomes quite big.

After the mind-map kind of converges, I need to serialize it and give the ideas an order, so that I can convert it into a text. The result is an outline. I now have a scaffold and a starting point.
I used two different paths with the mind-map and outline so far.
Producing a Text
I take the outline in markdown format, convert it the headings. Then for each heading, I start a bullet point list and write down everything in my head for the current section. As fast as possible. It is nice, that the very big task of producing an artifact is now a bunch of very easy, short and atomic parts. I do as many sections as I have energy and time and can interrupt and continue at any time. There are no dependencies. Even doing one is fine.
Then, I use a Language model to convert the awfully written bullet-points into a comprehensible text. LLMs are good in writing good enough text but have nothing to say by themselves - with the brain-dump, we give them something to say.
The resulting text is kind of bad, but it is a Draft. Not even the first Draft, thus “Draft 0”. Now I can rewrite each section on my own - section for section. Small, atomic tasks. I can do as many as I find energy to do without feeling lost in the process.
Producing a Presentation
Sometimes a spoken presentation is the better artifact for me. The process to get there is faster and more lightweight but similar. For each entry in the outline, I write the stuff I have to say about the topic as small bullet points. These are not all the content in my head, but reminder what I want to say.

The bullet-point list contains the outline and speaker notes. I let an LLM convert it into a format with heading, which pandoc understands. I use pandoc to create a pptx deck import it into Power Point, iterate on the slides, add images and make them ready to present. Then I give the presentation with minimal preparation. A lot of improvisation is required. But it is not too hard, as I have everything in my head what I want to say and the slides guide me.
If important, I can give the presentation multiple times and iterate on the content. I have never done it, but transcribing the final spoken version would be probably a nice first draft for a written text.
Speed and Ease
The goal is to make a complex task easy by breaking it into small atomic, interruptible tasks which I can do fast. I can do a presentation of one hour from before the mind-map within 45 minutes or so.