Today and Old School

Playing with an LLM is fun!

An LLM (Large Language Model) has to my model read all of human knowledge, and understood slightly.

There is no point where I have felt inferior to the model. The model performs what I expect, but saves the time and effort I would have spent. Keep in mind that I have compiled a considerable set of rules to guide LLMs in projects. All my life I have had a series of projects I would liked to have explored, but did not have time. Now suddenly, this is easy.

I spent much time in libraries, in the 1970s, chasing any question of interest. Later found I could make guess on where technology might go. Now I can set huge resource at work chasing any question I might want to explore.

This is fun!

I find using the present accessible version of AI (an LLM) much like building an auxiliary mind. That mind is (so far) slightly behind.

The upshot is well-researched, tested, generated code that I might have written.

Rather enjoying this. Current experiments in: https://woven1.bannister.us/ https://woven2.bannister.us/

Not sure where this leads, but sure it does.