> if it runs, it runs, regardless of who or what made it. If it fails to run who is responsible to fix ? Problem with AI is that it does not have causal-models

> if it runs, it runs, regardless of who or what made it. If it fails to run who is responsible to fix ? Problem with AI is that it does not have causal-models of how something works, and the reason for that is that it doesnt interact with the world. I think of it as an armchair expert who spouts recommendations without having a grounded understanding in the real world.