I agree that general consensus for traditional governance insists on human-in-the-loop or human-on-the-loop. Yet, I think it is an open question as to whether the same holds true in agentic environments. What do you think?
This is a very good question, Jason. Unfortunately, this is not the case in the agentic environment and I think that can be very dangerous. In very simple tasks and workflows, that could be ok. But, in critical situations, there should be a governance system in place in such a way that human should always be there before the output of one agent goes as the input of another. I define critical situations as any situation that materially affects humans such as medical insurance coverage, mortgages, healthcare, and the likes.
I agree that general consensus for traditional governance insists on human-in-the-loop or human-on-the-loop. Yet, I think it is an open question as to whether the same holds true in agentic environments. What do you think?
This is a very good question, Jason. Unfortunately, this is not the case in the agentic environment and I think that can be very dangerous. In very simple tasks and workflows, that could be ok. But, in critical situations, there should be a governance system in place in such a way that human should always be there before the output of one agent goes as the input of another. I define critical situations as any situation that materially affects humans such as medical insurance coverage, mortgages, healthcare, and the likes.