Core Theory

The basic idea Derivation relies on is that some knowledge can linked and connected further as well as more optimally than it currently is. It wants to assist in developing the processes of systematizing knowledge, acting like a backbone that makes different systems work better together. It would, further on, aim to function like a "global brain", connecting all sorts of knowledge networks, allowing people and their technology to connect in a decentralized way, to get more practical use out of it. When making complex conclusions, like in genetics, we often rely on a series of earlier findings from various fields, many of which were established long ago. As new observations are made and the world changes, these old conclusions need updating. Instead of doing this manually, flexible formulas could automatically update these conclusions using AI and machine learning.

Multiple Epistemologies

Throughout history and across different cultures, people have developed various ways of understanding and viewing knowledge, which can also be called different epistemologies. Modern natural science contains many of the most well-known. Still, there are also different epistemologies tied to religions (a Christian epistemology), philosophical and political ideologies (a Marxist epistemology), and others. People have sometimes believed that one way of knowing is better than others, leading to conflicts and destruction. Derivation aims to be open to all consistent methods of knowledge - those that don't contradict themselves, as Derivation cannot be of assistance to those. Since there are different types of knowledge, we need multiple systems to account for them. Often, different epistemologies lead to similar or overlapping results, creating common ground that can help bridge different networks of knowledge and even different worldviews. Examples of this could be that people of different religions or worldviews might disagree on a lot of things, but they probably agree of a lot of things as well, like for example, roughly, what water is, and that it's crucial for your health.

System Design and Features

The basic setup of a functional Derivation would be presented on a website and an app, with interconnected domains regularly updating each other. The site would be unconditionally accessible, with options to register for saving specific parts and customize the user interface. It will emphasize openness, impartiality, and transparency to avoid bias and prevent knowledge from being withheld for reasons of power, such as economic interests. The key features will include the ability to create, name, classify, and code arguments, sources, premises, and conclusions. Relevant sources will be linked and quoted, with the quotes interpreted by the user as premises. For security and relative reliability, sources should ideally be peer-reviewed, as well as in multiple different ways methodically scrutinized according to their epistemology, according to what is considered appropriate and relevant by its users. This setup will enable the creation of more arguments, potentially with the help of LLMs. Arguments and mathematical formulas will be displayed both formally and in plain language, as clearly as wanted by the user. The system would organize everything according to its epistemology and build knowledge networks that can be navigated like a search engine. Over time, it could lead to the creation of articles similar to an encyclopedia, with classifications inspired by systems like the Dewey Decimal Classification. The sources will be linked to, and not stored directly on the servers for Derivation, for at least quite a while. There would also be forums for discussing the structure, arguments, premises (similar to Talk pages on Wikipedia), and other relevant topics.

Scalability and Precision

The system would continously be growing based on the access of the resources that would allow it to do so. In what direction it would grow would be dependent on the practical use of it. As it's impossible to derive events and processes entirely, on a quantum level, it would always be a question of precision, the value of it, compared to the resources that are needed. With less precision, we need a higher margin of error. This is similar to how many decimals of pi we use in various operations. Epistemologies who are not easily systematized using logic might have issues with consistency, or the use of reliable methodologies. If they remain like that, they would not be rejected selectively, but rather find that Derivation might be unable to assist them. For clarity, it is useful to operationalize, stipulate and in turn derive the words used in an argument, as far as possible. Like with language technology.