These ideas are more like the unrealized gifts contained in books that sit on the shelf. In this case a literal bookshelf that has been reduced from the entire library of my college experience to just a little over 3 shelves. Some of the book covers have faded a little bit and others are the same as they were when they were purchased. Sometimes I wonder if I should spend more time with these books each day. Home libraries are not really a standard or even all that common anymore. I know the percentage of people buying books has dropped just like the percentage of people reading books. Even at school these days a lot of the content is on iPads or Chromebooks and physical books are far less common. This makes a lot of sense for schools, but it does somehow make me a little bit sad. Lockers full of books were a thing of the past that will be forgotten within a generation. Things change and that type of progress will continue as we move forward. 

I’m considering taking the time during this winter break to convert the weblog from an older theme called “Independent Publisher Publisher 2” to the standard “Twenty Twenty-Five” theme that was released as part of the WordPress 6.7 release. That might end up being a couple of design efforts, but it should be worth it to kickstart the new year with a new theme. At least I think the effort will be worth it. We will see what happens on that front as the actual work commences. I watched some videos about the Gutenberg design layouts and it seems to a lot more complex than my current setup. 

I have spent a lot of time thinking about the nature of before LLM writing and post LLM writing. All of the models training on content before the introduction of widely distributed artificial text are going to have a certain bias and content structure that is going to be changed by the post LLM writing period. Inherently within the process of creating this post LLM content the very nature of writing and creating has changed. I’m concerned that people will write differently after reading so much artificially generated content. That style will essentially be imprinted on the world post LLM creation. It will be a sea change in thought that cannot be reversed. Truly this watershed event will transform the output of the written word in ways that we are just beginning to understand. I don’t consider it to be a trend for the better. Granted that a lot of content being created was not at the peak of what was possible for humans to achieve. We may have hit the peak of quality writing output sometime in the last millennium. 

My biggest consideration for the day was to try to figure out if a few shot personal persona bot model creation from declaration or inference could work. My initial research questions would be pretty straightforward. 1) How many declarative statements about a person does it take to model a personal persona bot response pattern? 2) Do they have to come from the person or could they be inferred? 3) Would it be possible to set up some personal templates to make the few shot learning more effective?

Dr. Nels Lindahl
Broomfield, Colorado