Circling the Drain...Unlike Dorothy

Whether Wizard of Oz was really just a children’s story, or an argument in support of Populism, Greenbacks and Inflation, one fact remains true - Dorothy went around in circles. Several times, in fact. First, she was devoured by a twister, then she went to Oz only to set off on another journey, eventually returning to Oz. Finally, there was a great deal of turning around in circles and twisting toes to magically get transported back home. She came back to where she started - Kansas.

Yet, it is not the same. The difference was the journey, the adventure, the experience. Almost exactly one year ago, I found myself wide-eyed, jaws-dropped, leaning forward in my seat with fascination. Watching the Unsupervised Model placing words in clusters, each one closer to those with more common contexts, and wondering if a computer really could learn a language like us, my “gut” told me to chase it. My French teacher once told me that I had a “good gut”, instincts that rang true, feelings that itched for my attention. So, trusting my “gut”, I listened to my instincts, which clearly were screaming at me. My fascination, curiosity, wonder, all of which attracted me towards Machine Learning, were somehow “feel-good” emotions, unlike Gollum’s magical pull towards The Ring. No, this was simple curiosity, but it was still magic - partially because I had not the faintest idea about the intricate workings behind the neural networks.

This week, I realized I have made a circle, arriving back at what brought me to Machine Learning in the first place. I suspect this is one of several circles I will soon find myself twirling within. There is a sense of familiarity and peace when you make a circle, but I realized that more than a circle, this was a spiral. What was magical last year because of my ignorance, is now even more magical because of the bits and pieces of information I scraped together. To put it in one way, it is like enjoying music written in a different language, understanding the familiarity of it, yet not comprehending a single word.

What is rather concerning about this spiral I have begun swirling down is the drain at the bottom. While thinking about my Original Work Proposal, ideas twisted and turned in similar spirals, and a frightening thought occurred to me. Earlier this year, when I was developing web-applications and mobile-applications with Node.js and Swift, the amount of planning and design that went into the projects, accompanied by the amount of revision and cleaning-up shocked me. I started with a high-level design, then went into designing the classes, then their methods, simultaneously with the interactions within, creating what turned out to be a confusing, messy, tangled web. Taking each individual part of the design, I delved deeper, designing each individual component, untangling the knots, and retying them. I realized that I would have to go back into a similarly dense thicket once again, if not more thickety.

It seemed to me that each component was an infinitely spiralling drain on its own, and I worry that I will be stuck there, going in circles forever - designing, revising, coding, debugging, crying, sighing, screaming, revising, designing, coding, erasing, coding, deleting, designing, more sighing, more revising, more coding, more testing. At this rate, is there any light at the end of this...tunnel/drain?

These are worries for a later date. In fact, instead of viewing my travels as spirals, and falling down into a drain, I would like to see it as falling upwards - like Dorothy. I hope to view beautiful cities like Oz, only these will be multi-dimensional word vectors, data mined like emeralds. I hope to follow a yellow brick road - sparkling with my curiosity, glittering with guidance and rich treasures of knowledge. I hope to come back to last year in the future, knowing more about what I don’t know - coming back home to Kansas as a new, better, wiser me.