I think this chart explains the inherent contradiction at the heart of LLM liability: - Management demands more productivity (which let's say comes from using LLMs) - At the same time blames the human

I think this chart explains the inherent contradiction at the heart of LLM liability: - Management demands more productivity (which let's say comes from using LLMs) - At the same time blames the human for any mistakes LLMs are inherently unreliable, by design. You can't have it both ways, either it's fast and reliable or slow and supervised.