THE ROBOT AND THE BABY
What Does The Future Hold for AIs Now?
Note: Grown-up language is used in the linked short story.
Following up on the theme of a post by JoJo Stratton earlier, and in consideration of the potential results of #Recursion , the linked story came to mind. Its author, computer/cognitive scientist John McCarthy, penned it a decade ago, seven years before his death. (Scroll down, it's a scanned/OCR'd archive so there's odd whitespace.)
Intended to be provocative, it touches on ideas first popularized by Isaac Asimov, and discussed widely in the last seventy years under the general theme of "Laws of Robotics". ( http://en.wikipedia.org/wiki/Laws_of_robotics )
Are there ethical limits to the capabilities we should allow AIs to acquire, or ways to ensure they are acquired only with these kinds of limitations on their use? Is it even possible to "build in" such limitations, as Asimov described? Finally, as the story forces us to ask: Will greed and/or politics sabotage ethical intentions?
http://www-formal.stanford.edu/jmc/robotandbaby/robotandbaby.html
http://www-formal.stanford.edu/jmc/robotandbaby/robotandbaby.html
Thanks for the reference, Jon. I'll add it to my reading list.
ReplyDeleteThat's a really neat short story; thanks for sharing it.
ReplyDeleteWow! That was a great read! To your last question: human nature (base emotions, greed, anger, love, and yes, politics ;)) I think will always get in the way of ethics. Greed is present even in the littlest children, everyone wants the same toy and 'sharing' is a learned concept. Ethics are almost an idealistic notion of the 'perfect world'...and in anyone's version of that perfect world, someone else has to give up something that is precious to them (I.e. freedom, mental or otherwise). It would be hotly contested for which ideas (or who's ideas) would be employed in such programming, and someone always loses. Even 'do not kill humans' could turn into infinite human captivity with bread and water to protect us from ourselves.
ReplyDeletealso adding to my list, then will have questions : ) Jon Luning and others too, to really look at the concepts associated with ADA what other items should be on a reading list?? Cherie Brush Cindy Woodman Michelle E
ReplyDeleteCindy Woodman I dont know that sharing is a learned concept per sae. Empathy comes naturally, just not nessisarily right away... There was a study with rats where a rat exposed to another rat restricted in a very tight cage. When the trapped rat was freed, the rat that wasnt trapped shares treats with the previously trapped rat. But Im not a developmental psychiatrist. So the diffrence between not having yet developed empathy yet due to age and just not having it or not caring isnt something I know a lot about. I do know a developmental psychiatrist. I could ask him for some light reading if people are interested in that.
ReplyDeleteMichelle E maybe it's only the adult human's idea of what sharing should be that people feel the need to impress upon their children? I'm definitely not a developmental psychiatrist...but now that you mention it, I bet they could shed a little light on the idea of what a highly logical, brilliant mind might need to do to learn emotions...
ReplyDeleteJoJo Stratton Mostly so far I've been wandering more into Wikipedia via the links I already provided. I'm sure I'll run into more relevant info to share, though, as I start looking into things from the technical side, as I describe in the "novel" I just posted. (Me and "brief" just don't go well together. ;-))
ReplyDeleteIf you want to go waaay down the rabbit hole, JoJo Stratton , the book that caused me to reformulate my concepts about "intelligence" (and several other things) was Gödel, Escher, Bach: An Eternal Golden Braid by Douglas Hofstadter. It was recommended to me by, and then borrowed from, a friend a few years after it came out. (I believe it won the Pulitzer Prize in General Nonfiction in 1980.) I bought that friend a replacement copy five years later, when I had thoroughly dog-eared the original and had not yet "completed" the book. In hindsight, it probably took me about 8 years to read, as I would often stop and need to think about things. For a week. Or more. For many people I know, it answered the deeply puzzling question of "How could machines possibly 'think'?"
ReplyDeleteFor a shorter, somewhat lighter, and very interesting read, I'd recommend the book Hofstadter co-"arranged" with Daniel Dennett in 1981: The Mind's I - Fantasies and Reflections On Self and Soul. It includes 27 short stories, each followed with some interesting musings. It was in this collection that I first read "The Soul of the Mark III Beast", which I linked in the comments of an earlier post.