I think I'm one of the slopsters, amongst other things trying to create accurate measurements of harm to unarmed people around the globe. Then a enablement matrix, so it maps culpability. Expected harm if continued and life giving measures proposals. No coding experience other than dabbling over a few months. Predictive modeling maths, is beyond me, want an accepted calibrated method that is uncontroversial but seeks to give a true reflection. Many measurements are contested and reality gets missed.

Thanks, a great read, as a non programmer, playing around, I find the conceptualisation of ideas of llm models great and pulling up some research. As a pauper, it has made idea progression better though exquisite control takes time and often get caught in loops. I was beginning to find the limitations and that it was quicker in some instances to mix it up. As a stulator and drafting it's good. Finished article, long form content, quite poor.

Must read, enjoying the articles folk are writing