Totally agree on the smaller scale local models. But this is my exact point here, it’s not inherently evil, it’s the implementation and systemic issues that are the issue. Yet people act like all AI is inherently evil and wrong.