TSR #10: Creative Processes

post by Hazard · 2018-01-17T03:05:18.903Z · LW · GW · 0 comments

Contents

No comments

This is part of a series of posts where I call out some ideas from the latest edition of The Strategic Review (written by Sebastian Marshall), and give some prompts and questions that I think people might find useful to answer. I include a summary of the most recent edition, but it's not a replacement for reading the actual article. Sebastian is an excellent writer, and your life will be full of sadness if you don't read his piece. The link is below.

Background Ops #10: Creative Processes

SUMMARY

*Just realised that there were two more babble posts, and what I point out as anchors and what alkjash called beacons

I’ve really enjoyed the ideas and discussion in the Babble post and its follow up, and will be looking at this TSR edition through the lens of babble and prune.

The standout claim in this article is that there are ways to structure your approach to a creative task such that you don’t end up stuck. What’s going on when you get stuck? Getting stuck is when your pruning filter has a higher standard than the quality of your babble, leaving you with “nothing to say” and an unsatisfying feeling of being stuck.

What then, would be the solution to getting stuck? As people were pointing out, often the first step is to tone down the aggressiveness of your pruning. Babbling nonsense beats not saying anything. But what comes after that? How do you end up getting your babble to a point where even with whatever pruning standard you think is appropriate, you still end up producing at that level?

Thinking of your babble generator as traversing through a connected babble/idea graph, one route seems to be to improve the idea graph itself. Learn cooler things, and make more connections between the cool things. Another approach is to train the part of the generator that makes the traversal decisions. The traversing part of the algorithm seems to include a mechanic for deciding where to restart from for the next traversal, and how to choose which neighboring node to visit next. Improving the graph, or improving the traversal would produce better babble.

The writing process that Sebastian draws attention to a specific way to improve your traversal quality, within the context of your current writing project.

It’s non-controversial to claim that knowing what you are writing about makes it easier to write. Why, from a babble and prune lens, is that so?Well it seems like your traversal algorithm is affected both by some “general ability” and by whatever is in your working memory. What is on you mind anchors how you traverse your babble graph.

If the essence of what you want to say is already in your working memory, you can produce babble that matches said essence much more easily. Writing an outline is just a way of expanding your working memory, and if you make a quality outline, it will have a strong anchoring effect on your babble. The cool thing here is that you can get a huge context-specific boost in babble quality, without the work needed to improve your babble graph itself (though I’m sure it’s possible to have a poorly enough connected babble graph that no anchor will help you that much).

So how do you make the outline in the first place? Recursively using babble and prune! Think back to the 5 stages of writing that Sebastian mentions. Brainstorming looks like pure babble. Categorizing looks like babble anchored by your previous babble, with a lite amount of pruning. All of that babble is used to anchor the babble and prune which goes towards making the outline. If you make a good enough outline, you can probably babble your way through the entire writing phrase, and just swing back around for one last editorial prune to get things ship shape.

In the most general sense, you might be able to describe a generic creative process as “Recursively using babble and prune with progressively strong pruning to create anchors of high enough quality to allow you to babble to main content at the standards of your full throttle prune.”

For writing in particular, I’d guess it would be more useful to follow a specific five steps like Sebastian laid out, as opposed to thinking in terms of the general principle. Though it does seem like the general principle is a good anchor to have in your working memory for when you are trying to come up with a flow for your specific creative work.

Some questions that could be useful to answer:

0 comments

Comments sorted by top scores.