X-Message-Number: 16186 Subject: Re: Important question for the isomorphists Date: Thu, 03 May 2001 16:25:32 -0500 From: Will Dye <> In response to a statement that a machine-based life form is obviously not a "person", Jeffrey Soreff replied: > As far as I can tell, a sufficiently good simulation of a > person would have the same subjective experience and the > same social interactions as a biological person. I, for > one, would include such simulations in the same category > of persons as the biological ones. [Caveat: I have not followed CryoNet lately, so I'm jumping into this thread mid-stream, and only then because I noticed Dr. Soreff's name on a posting.] I don't think we're going to settle the issue of "what is the precise definition of an entity that merits legal protection" anytime soon. Even so, I guess Jeff is right that we should still put in a word of defense now and then when someone says something particularly disturbing. Consider this to be a me-too post, supporting Jeff's defense of machine life forms. I agree that we should not, at this early stage of the game, make broad statements that such entities are not "persons". As technology continues to develop, we're going to get more and more of these "where do you draw the line" issues, if only because technology will give us more ways to create stuff that has no clear historical precedent. I have a lot to say about general approaches to such problems, but for now I'll offer what I think is an important take-home thought: Relax a little. Apart from regulatory considerations, the technologies which make it difficult to draw the lines between life and non-life also tend to make it less costly to draw those lines generously -- giving the benefit of the doubt to the borderline entity, without imposing an unbearable burden on any other entities. In this case: if we have the technological means to create full-blown machine intelligence, we should also have a much greater ability to provide all of the entities involved with whatever resources they need in order to survive and prosper. There's no strong need to start declaring broad categories of entities to be "obviously" unfit for human respect and love. Obviously, this one principle won't solve every problem. What some entites want is to destroy all other entities, so there's no pleasing everyone. :-) We still have to fret about safety issues, so the delineation problem cannot be avoided outright. Humanitarian issues, among others, are also important enough to merit wading into the murky multidimensional waters. Even so, this principle might be a source of reassurance to those who get frustrated by the inherent difficulty of the general delineation problem. The forces which tend to make precise delineation more difficult will usually also make precise delineation less critical, at least in terms of basic needs. We can more easily afford to worry less about parsing the definition of "neighbor", and focus more on practicing the kind of behavior that makes neighborliness worthwhile. So please, let's give the bots a break. --Will Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=16186