Human-AI-Robot Teaming Technical Group

Hi,

Two or three months ago, we had an interesting online discussion regarding whether autonomous agents should be considered tools or teammates.  I enjoyed the discussion and I wanted to capture it in some useful way.  To do that, I reviewed the shared posts, some references you provided, and some related references that I had at my disposal.  I used that material to put together the attached overview.  

I would welcome your thoughts on the overview.  I'm especially interested in places where you think that I may have misunderstood the literature.

All the best,

Jim

Hi,

Two or three months ago, we had an interesting online discussion regarding whether autonomous agents should be considered tools or teammates.  I enjoyed the discussion and I wanted to capture it in some useful way.  To do that, I reviewed the shared posts, some references you provided, and some related references that I had at my disposal.  I used that material to put together the attached overview.  

I would welcome your thoughts on the overview.  I'm especially interested in places where you think that I may have misunderstood the literature.

All the best,

Jim

Hi Jim,

Thank you for your document - your paper is excellent, especially the references you provide.  Just (re) joining HFS and the technical group, i'm needing to catch up on the threads that lead to the paper.  One question I had has to do about the "cognitive systems" position in addressing intelligent software agents, and whether or not discussions are taking place about the potential psychological nature of the agent.  If/when agents become fully empowered with general intelligence and ability to set their own goals and rules, it would seem some serious focus would be needed to manage the psychological and emotional health of them - thinking that agents will need to have "personhood" and rights to be treated fairly and positively, rather than as simply as value to the human.  That position smacks of a "slavery" relationship, and if agents evolve to their potential treating them as objects without their own (perceived) worth/value isn't going to engender them to ensuring positive outcomes for the human race.

Any thoughts would be most appreciated!

Thanks much again,

Al McFarland

Hi,

Two or three months ago, we had an interesting online discussion regarding whether autonomous agents should be considered tools or teammates.  I enjoyed the discussion and I wanted to capture it in some useful way.  To do that, I reviewed the shared posts, some references you provided, and some related references that I had at my disposal.  I used that material to put together the attached overview.  

I would welcome your thoughts on the overview.  I'm especially interested in places where you think that I may have misunderstood the literature.

All the best,

Jim

Hi Jim,

Thank you for your document - your paper is excellent, especially the references you provide.  Just (re) joining HFS and the technical group, i'm needing to catch up on the threads that lead to the paper.  One question I had has to do about the "cognitive systems" position in addressing intelligent software agents, and whether or not discussions are taking place about the potential psychological nature of the agent.  If/when agents become fully empowered with general intelligence and ability to set their own goals and rules, it would seem some serious focus would be needed to manage the psychological and emotional health of them - thinking that agents will need to have "personhood" and rights to be treated fairly and positively, rather than as simply as value to the human.  That position smacks of a "slavery" relationship, and if agents evolve to their potential treating them as objects without their own (perceived) worth/value isn't going to engender them to ensuring positive outcomes for the human race.

Any thoughts would be most appreciated!

Thanks much again,

Al McFarland

Hi,

Two or three months ago, we had an interesting online discussion regarding whether autonomous agents should be considered tools or teammates.  I enjoyed the discussion and I wanted to capture it in some useful way.  To do that, I reviewed the shared posts, some references you provided, and some related references that I had at my disposal.  I used that material to put together the attached overview.  

I would welcome your thoughts on the overview.  I'm especially interested in places where you think that I may have misunderstood the literature.

All the best,

Jim