Things are moving along a little slowly today. Getting back into the swing of daily writing has required a little more focus than it used to at the start of a daily writing session. Strangely enough my Chromebook just tried to pair with my Pixel Buds which was awkward when my phone started blaring songs from Pandora internet radio. Instead of trying to deal with that problem I decided to just turn off Bluetooth on the Chromebook. That is not something I need to write this weblog post. My audio has now been moved back over to the right source. Things are starting to move along. My thoughts are now really focused on the task at hand. Being able to find that moment of focus is an important skill for a writer to have developed and practiced.
Yesterday I started reading an article from Nature published back in 1976 by Robert May on chaos theory. The PDF I came across was only 9 pages. Two real methods exist to absorb a new academic subject. Probably the majority of the time people learn a new subject from a textbook based learning approach. Outside of that you can read primary articles about that subject. Generally as you start the academic experience you end up moving from textbook based learning to article based reading around the time you move to graduate school. Currently I read a lot more academic articles in PDF form than textbooks. That is a function of cost and availability. Within the field of machine learning and artificial intelligence people are writing textbooks and sharing them as open source documents. You certainly can read them in that form.
At the moment the degree of change within the field is so accelerated that I mostly end up reading articles in PDF form as I chase the bleeding edge of the discipline. Every weekend and during the course of a lot of weekdays I share links to them and dig into the various ways people are pushing the discipline forward. Honestly, the sheer volume of content being produced would overwhelm any reader or entrant into this field of study. More papers and content than could possibly be consumed are coming out in a variety of ways. Knowledge is no longer being gated by the reviewer based framework of the best academic journals. People have just elected to share pre-prints and sometimes they don’t even try to get those articles published. They share them and move on to the next article writing project. We have within that framework a possibility for extreme output. Within the ML/AI space that is certainly the case right now.
Understanding things has really required a lot of study. I have written 75 Substack posts about topics in the field of study and have plans to finish a package of 104 total foundational essays on the subject. You can imagine anything deep enough to support that many essays would be vast and rapidly changing. That for sure is the case. I’m going to keep going and learning and striving to learn as much as I possibly can to share with people and to advance my own individual learning.