Many of the arguments about LLMs seem to involve us talking past one another. Insistence that they are “just autocomplete” is demonstrably true, but often remains too abstract to persuade people. I have tried to be less abstract here. Meanwhile, most proponents at some point break down in frustration and say “just try it and you’ll see!” Their argument is phenomenological: Doesn’t it feel smart and capable? Don’t you feel like you’re getting value out of it? Working faster? This, too, is demonstrably true. Many people feel that way. The problem comes when we mistake the feeling of using a chatbot (writing this input and getting that output feels like talking to an intelligent person) for the actual inner mechanisms of it (next word prediction). As with calculators, many different internal mechanisms can produce indistinguishable output.

Infinite* Slop

Infinite* Slop is an enterprise-ready high-performance slop generation solution, designed to waste resources of shitty web crawlers and potentially ruin training sets of unethically-sourced AI projects.

I'm too old-fashioned to use Markov chains so this just uses the good old template-based random string generation approach from the 90's.

Artificial intelligence (AI) technologies are being embedded into everyday life by powerful actors, primarily motivated by profit. Police, border and criminal justice agencies are also looking to take advantage of the new powers AI offers for “security” policies, at both national and EU level. The EU is creating new infrastructure, away from the public eye, to allow the swift development and deployment of “security AI.” This will also reinforce the existing discrimination, violence and harm caused by policing, border and criminal justice policies. Exposing and understanding this emerging security AI complex is the first step to challenging it.