I’m on a flight to Orlando, Florida right now. We departed from Denver, Colorado, after apparently some issue with a tray table was fixed. As delays go this one was not a very big deal and only amounted to around 30 extra minutes of standing. I’m pretty confident that we will make up a bit of time during the actual flight. We appear to be traveling at roughly 558 miles per hour. 

I’m going to spend this 3 hour block of time just writing. That is about to happen. You certainly could sit back, relax, and prepare to read a bunch of prose. 

During the course of cleaning out a drawer and a stack of papers I briefly considered going down the path of working on some public administration related academic articles. That type of consideration happens from time to time and is one of the things that at some point will probably happen. My 5 year writing trajectory includes a lot of things to complete. One of the things that has been a major part of that effort for the last 2 years has been working on Substack posts. That is a pretty distinct weekly cycle of creation to publish. 

What I plan on doing here after hitting the two year mark on that one is working toward a different method of weekly creation. Each weekend I’m going to work on creating academic articles and some of that content will be used to feed the Substack posting. Over the last two years things happened a bit differently. The content being created for Substack sometimes found its way into an article or manuscript. Flipping that scenario around will be key to boosting my productivity in terms of academic prose output. 

That will for the most part give me a set of two windows of 2-3 hours for crafting and generating academic content each weekend. Over time that will be enough to build out and write a bunch of different things. Some of that for sure will be literature reviews. Another part of it will have to be using a combination of publicly accessible dates and automated sentiment analysis to create original work. I should be able to build a sentiment analysis model automation and execute it for each academic paper if that is the dirrention things end up going. It will certainly yield original content. 

I think as we work into 2023 that is one of the things that I’m most looking forward to producing. Within the machine learning space I have produced original research into the MLOps space and looked at a lot of open source code repositories in that space. At this point, I want to shift away from searching out and reading academic articles and move from conducting an ongoing literature review in the machine learning space to producing some content that makes a substantial contribution. The field has had so much content added recently that just sorting out the ideas coming in is hard enough without trying to consume the best of it or even locate the very best of it. This has to be the peak academic period for machine learning based on the hype cycle. I think at some point in the next couple of years a lot of machine learning researchers will be shifting over to the study of artificial intelligence in general. 

That shift is probably good overall. My views which have been shared previously are that a lot of both model production and intellectual overcrowding has occurred within the machine learning space. A lot of papers have been written. It’s great that people are producing content and I’m happy they are mostly sharing it online without any paywalls or other limitations. You will have to admit that the sheer volume of it is overwhelming and at this point beyond any method of conventional sorting. Some of it gets promoted by social media and other actual interpersonal interactions where content is shared. People are having to really go out and share a paper to get it noticed or picked up by the top contributors within the field. 

I think at this point in the program I’m going to drop out of weblog writing mode and work on some different essay related content.  I might circle back to this one or something else might show up.