After writing a few blog posts, I thought it’d be appropriate to write a somewhat meta blog post regarding my experience with blogging so far. My schedule tightened up towards the end of the semester, which left me unable to continue my (until then unintentionally regular) Saturday blogging ritual. Meanwhile, I’d been thinking about a few things, including:

• How to choose topics to write about?
• Are there cognitive benefits to writing?
• Where should I be sharing my writings, if at all?

To summarize before diving into greater detail, I’ve thoroughly enjoyed the process of writing these posts, likely because they’ve allowed me to articulate my thoughts and get my mind thinking. However, I wonder if it would help to approach writing and topic selection with a bit more purpose (e.g. write to inform a broader audience about technical topics) rather than just writing down what I’m thinking at the moment. I’m a little afraid that my current approach is motivated more by my enjoyment of the actual process of writing rather than by the content that’s produced, but perhaps that’s okay if written in a different format - a journal, for instance. I’ll first talk a bit about my current workflow and then elaborate on the so-called “cognitive effects” that I’ve been experiencing.

My recent blog posts have touched upon random ideas I’ve had about metareasoning, the relation between the trustworthiness and fakability of media, and effect of public perception of science on science itself. The process of writing each post has usually been as follows. I spend part of the week thinking abstractly about some topics of interest, often when some kind of event of interest emerges; for example, in response to NVIDIA’s fake celebrity GAN paper, I had a starting point for a discussion about digital media. Then, on Saturday, I spend the morning and early afternoon doing any necessary research and writing most of the post. I then do a quick pass on Sunday morning and hit publish. I was able to keep this up for three weeks in a row and each time, I noticed an interesting (you could say) “cognitive” development as I was reading papers, writing code, or chatting casually with others, etc. It felt almost as if my brain had been given a 25% boost, enabling me to process complex ideas and ask thoughtful questions with much greater ease than I was used to; I suppose this could also be referred to as “dusting off the cobwebs” or “clearing the ‘brain fog’”. But this temporary effect seemed to fall off throughout the week, at least until I wrote the next blog post. I’ve been trying to determine how much of this is due to the act of writing itself (currently, I write fairly little from day to day) as opposed to the process of researching and thinking about some particular topic.

It would seem that the actual process of writing by itself only has some contribution to this effect. I’ve felt traces of the same effect when writing longer form messages and technical documentation, but only at most “¼” of what I’ve gotten out of these blog posts. I thus believe that the content as well as its presentation and/or style must have some bearing on this. These blog posts are decently long but are generally of the sort where an idea has to phrased properly in order to be expressed properly, whereas comparatively little thought goes into instant messaging chats and straightforward technical documentation (see this for an example of hastily written documentation). I generally find myself writing and rewriting each sentence two or three times at least, so perhaps that iterative and thoughtful process is a contributing factor. I also speculate that the process of taking in a potentially disconnected body of knowledge and synthesizing my own insights helps to stimulate my creativity. Lastly, I wonder about the role of the blog’s public nature (and associated pressure + fulfillment from audience engagement). I could just as easily record all of these as private journal entries but the fact that these could be read by someone other than myself greatly incentivises clarity.

Is there some way of choosing topics that (1) satisfies the criteria above, (2) allows me to learn and reason about new, interesting topics, and (3) is a worthwhile read for others? I can easily see how an excessive desire to satisfy (3) can take away from the creativity and openness one might want from blogging, but sometimes having that external validation can be the primary force that sustains the blog in the first place. It might help to choose a theme. I’m quite fond of the kinds of posts on distill.pub, so perhaps I could go down the path of explaining machine learning papers that I find interesting and/or creating interactive visualizations. These themed blogs could be organized by category (e.g. Self-reflection/introspection, technical/research-related pieces, philosophy). I could also move some of the day-to-day reflections into a journal if only to reduce the noise in the publicly viewable writings. I also wonder that as I become more involved with the process of writing a research paper if I’ll get a similar sense of fulfillment and mental clarity.

I’ll finish up by briefly reasoning about the process of writing this particular blog post. The content consisted largely of my ongoing speculations rather than more fully formed opinions, so it would seem that the lack of clear structure made it easier to write but not quite as “fulfilling”. It was still quite helpful to go through the process of writing the post. I feel that I now have a clearer sense of the contributing factors to the “cognitive” phenomenon that I was observing and more direction on how to proceed with this blog on the future. Over the coming months, I’ll continue to experiment with various writing outlets and see how they compare. Overall, it’s already been a fun and fulfilling process so far and I look forward to seeing how it pans out in the long term.