This is me documenting a journey into understanding the LLM universe. I’m very skeptical of the cheerleaders, and since it would seem LLMs are not going anywhere soon, it’s time to learn. I started out by looking up how to do a few simple tasks, and a huge amount of the stuff out there doesn’t work at all (probably written by the LLMs).

These posts document my journey, and will hopefully be useful to someone else. Fundamentally these are my notes for when I eventually screw something up so bad I have to start over.

It will all be written by me. I’m very much of the opinion the writing from an LLM is flat and boring. If you think AI is a good writer, you’re a bad writer.

If you want to complain, please Contact Me

FAQ#

I am preparing a FAQ to answer some questions you might have

Nothing here makes sense, why not read tutorials and follow those?#

This is how I learn things. I need to fail a lot first so I can figure out what I don’t know, then I look at what the smart people are doing.

Why not just use OpenAI or Anthropic?#

Because I hope to do things would end up as multi thousand dollars of cost. My long term goal is to write some agents to do some long running security tasks 24/7.

But local LLMs are worse than the big LLM services!#

I for one am shocked a company that loses more money in a quarter than most US states have for yearly budgets would have better performance than my $2000 computer!

Nothing you do makes any sense, are you some sort of idiot?#

Almost certainly, yes.

Do you think LLMs will give us AGI?#

There is zero evidence we should expect this to happen. Part of the job of a skeptic is to use evidence to come to conclusions. There is no evidence for this argument. I am reminded of a scene from the Simpsons