Skip to main content

· 2 min read
Vinnie

When working with tools/ideas, it's important to understand how the layers of abstraction are stacked. LLM is as good as the layers underneath. Tools constantly change, but the electrical base subsystem is the same. Electron movement. You can also build "false floors" or layers of abstraction that follow circular logic or in direct contradiction to the cause/effect outcome.

When it comes to learning, it's best to learn from the ground up but apply with the highest level of abstraction to create the most efficient products. With so many abstracted tools out there to build off, the most important component is the foundation through the spine (the stacK). It's the reason why the Hypervisor --> VM relationship was a fantastic model until containerization became dominant for long term scalability, rapid deployment and better I/O functionality. It's the expand and contract framework that is forever in duality when it comes to development. It's also why the best front end is no front end. It meets the MVP idealism and efficient design, albeit hard to grasp.

This goes to conceptual building blocks. 3 main components, mathematics (in which all frameworks are built), history (in which all knowledge is based) and ? (in which we are unaware and therefore ignorant). As logic is applied mathematics, and code is a series of logic gates and LLM is based upon the sum of the knowledge obtained, it is important to create your model based on the best logic and the most amount of history.

Conceptually based upon this, censorship and misinformation are based on illogical agendas that should never be enforced, as while going back to the root on either argument will be based upon hollow layers of abstraction.