The Quixotic Divide
Why do we need to divide knowledge?
Welcome to Part Four of my discussion of the Holordiyu categorization system at the heart of the Manatee Search Engine proposal.
Even if Web searches aren't your area of interest, stick around. There's more to this idea than just improving search engines. It could change the way we discuss everything.
As I've mentioned in https://lincolnsayger.com/ccow/about.php one of the major problems with search engines today is that they don't give us what we want. While this is largely the result of corporations prioritizing profit over the well-being of users, there are technical limitations to search algorithms that campy be overcome solely through better natural language models, no matter how much artificial intelligence and money are thrown at the problem.
Large language models aren't capable of understanding context, and they suffer from a bias toward commonality that virtually guarantees that you won't find anything esoteric using their guidance. Only what is most common is most likely to be regurgitated, and what is most common on the Web is likely to be inaccurate or irrelevant to a search for something specific.
In addition, site owners have an incentive to deceive people to drive traffic to their publications.
So, we see three major problems (among others) with the search engines that have the best intentions: they don't serve rate but accurate results, they can't adequately prevent unrelated sites from coopting popular search terms, and they can't tell the difference between Shalimar, FL, the Shalimar section of the Kolkata Railway, the perfume by the same name, and a band named Shalimar.
Enter the card catalog of the Web, and effort to allow does to be categorized by directories and/or their publishers in a way that is structured semantically (diminishing the difficulties of foreign languages) and discourages cheating.
Beyond that, however, there is a motive here of improving philosophical and technical discourse about various topics.
And this system can also inform your use of KMSes, including as an extension to systems like Zettelkasten or the idea compass.
This system gives us a structured vocabulary with which to examine the aspects of an idea, phenomenon, person, or general topic. What is the core of the subject? What is the history of the subject? What techniques and tools are informed by or suggested from the subject? How is this subject converted from place to place? How does our thinking about it affect our world and behavior? What is the philosophy of the subject? What aspects of the subject are unseen or unexplored? What is the aesthetic of the subject? Are there individual variations for this subject? How much variety and uniqueness exists? What is forbidden by, to, or about this subject?
The model raises all of these questions and gives us a framework to discuss them. And it is flexible enough to work with science, politics, economics, religion, art, sociology, and history, among many others.
Using this could benefit society greatly, and building it into our Web landscape could improve research by an order of magnitude in quality and efficiency. So, I sincerely hope that you'll get involved by commenting on the articles in this series, to improve on the thinking I've put into it, and by using these aspects in your own work to help others find your writings. Think how much better your experience would be, if you could filter the results before they were compiled for your query, to only those sites that were relevant.
If you missed Part One, you might want to look back at it, now that you've seen my reasoning behind this series.
Please let me know what you think. I'd love to hear your reactions.