I’m a philosopher by training and by heart, as well as a software engineer by practice and also by heart. Recently, I found myself stuck in my career, which seemed like the right time to consider my next steps. I have a deep love for mysteries and abstraction, and there’s not much more mysterious and abstract than deep learning. It feels almost like unraveling the mystery of a thinking being. Yeah, yeah—I know they aren’t conscious, not even close… but how much is needed to awaken consciousness once all the fundamentals are in place?

So, I decided to pursue new training to familiarize myself with all this AI stuff. My approach is to dive as deep under the hood as possible while still keeping an eye on the whole picture of the machine learning system as a system. I aim not to get too nitpicky about how slight tweaks in equations can optimize results, or how tiny adjustments in algorithms can speed up the learning process or prediction speed.

Here are my hallucinations, since I’m too fresh in this field to call them anything else. It’s just a bunch of fresh ideas, a log of my journey, and impressions. This isn’t guide for newcomers – it’s experience of a learner.