I’m fairly pleased with how my posts exploring MLflow turned out. That said, I thought I would finish part 3 much quicker than I actually did. I took 19 days between posting the first drafts of part 2 and of part 3. Neither the coding exercise nor the analysis took remotely that amount of time.
I wonder if one contributing factor relates to the fact that this was a series of posts. I knew in advance that I wanted to split this topic into three separate posts, and roughly where each would focus, but I didn’t actually work on one part until I published the previous part. I have a suspicion that this is relevant:
Many years ago, I remember reading some claim that declaring what you want to accomplish (attaining some milestone, forming some habit, etc.) to other people actually reduces your likelihood of success. With some light research, I traced it to a psychology paper from Gollwitzer, Sheeran, Michalski and Seifert (2009), also popularized in a TED talk from 2010 titled “Keep your goals to yourself”. I thought I recalled that the mechanism behind this claim is that by telling other people of your intentions, you visualize and act out something resembling completing the goal, which triggers an endorphin rush similar to one from actually succeeding at the goal and that this satisfaction reduces your motivation to continue on. However, the original study’s proposal was more nuanced than this. As summarized in a contemporary article by Art Markman in Psychology Today:
Identity goals are goals that ultimately influence a person’s concept of who they are… [Gollwitzer, Sheeran, Michalski and Seifert] suggest that when people announce an intention to commit to an identity goal in public, that announcement may actually backfire. Imagine, for example, that Mary wants to become a psychologist. She tells Herb that she wants to pursue this career and that she is going to study hard in her classes. However, just by telling Herb her intention, she knows that Herb is already starting to think of her as a psychologist. So, she has achieved part of her identity goal just by telling Herb about it. Oddly enough, that can actually decrease the likelihood that Mary will study hard.
A followup study published in 2020 further suggests that telling others about your intentions may increase your chances of success, but only if you consider your audience “higher status” than you. Seems plausible, but vaguely patronizing.
My point is - I suspect that when I had completed two of the three parts in the blog post series, on some subconscious level I felt less urgency in wrapping up the third one. It’s a convenient story, but difficult for me to rigorously test. Instead, I’ve decided that for the next several multipart post series, I’ll finish all of the prep work (if not the writing) before I publish the first part live. Let’s see if that makes it easier to keep up momentum to the end.
Wait… did I just undermine myself by writing this out?!