I like the image of an orchestra of agents - and agree on the importance of AI systems to check on other AI systems. Unfortunately, the idea of a "human in the loop" in popular debate is often seen as a human checking in on an AI system rather than a human checking on alerts that a controlling agent sends to the supervisor. A lot of education to be done here!
For sure! I also think that the human in the loop has gotten too entrenched. It is lazy thinking, suggesting that just because you are in the loop the results will be better - when in fact we know that humans often defer to the machine when the machine is least accurate.
I like the image of an orchestra of agents - and agree on the importance of AI systems to check on other AI systems. Unfortunately, the idea of a "human in the loop" in popular debate is often seen as a human checking in on an AI system rather than a human checking on alerts that a controlling agent sends to the supervisor. A lot of education to be done here!
For sure! I also think that the human in the loop has gotten too entrenched. It is lazy thinking, suggesting that just because you are in the loop the results will be better - when in fact we know that humans often defer to the machine when the machine is least accurate.