Article View: comp.robotics.misc
Article #137Re: Emotion in Robots
From: Arthur T. Murray
Date: Mon, 20 Dec 1999 00:00
Date: Mon, 20 Dec 1999 00:00
62 lines
3791 bytes
3791 bytes
Gordon McComb, gmccomb@gmccomb.com, wrote on Sun, 19 Dec 1999: > JimT9999 wrote: >> Pain & Fear >> Keep in mind that all bumper switches are now considered "painful." > <rest snipped> GMcC: > How do you reconcile these "emotions" in a robot designed > specifically for making physical contact? If a robot is > made to sense its environment via tactile feedback, how are > these senses then considered painful? ATM: http://www.geocities.com/Athens/Agora/7256/emotion.html is a sub- page on how eventually to code emotion into Mind.Forth Robot AI. A sensor for tactile feedback will only register *pain* if signals from the sensor enter the mindgrid and activate special grids that are *interpreted* as pain: burning, or impact, or freezing, etc. It is theorized that physical pain and pleasure are not emotions, but are instead *translations* of specially dedicated neural signals into the *qualia* -- the phenomena -- of physical pain and pleasure. Personally I speculate that pain is evinced in a mindgrid when the interleaved neurons of the "quale" of pain fire at a faster frequency than the rest of the mindgrid, literally forcing the consciousness to be unshakeably aware of the cause of the pain. Pleasure, by the same token, is a slowing down of the rate of firing of the pain-pleasure intergrid, so that the rest of the mind diverts its attention from everyday concerns and notices -- because pain and pleasure are functions of *attention* -- that a pleasurable sensation is saying, "Whoa! Savor the moment." Thus to slow reality down is pleasurable, but to speed it up is painful. Most important here is the *translation* of signals. Why should a faster signal be painful? Because we don't want to pay attention to a burn wound, or a cut, but we are forced to. Our consciousness tries to escape but cannot, so it feels pain. When consciousness indulges in a slow signal, it feels pleasure. GMcC: > It seems to me that such > emotion is then dependent on the design of the robot. But isn't > one of the chief values of emotions is that they are, to a great > extent, consistent (therefore they can be given understandable > terms we can all use so we know what we're talking about)? > For humans, touch can bring both pleasure and pain, not only > by degree, but by association with what was touched. > I understand there has been quite a bit of study and discussion > regarding a "light seeking robot" exhibiting either love, or > aggression, or maybe both. Personally, I think it's all hokum. > I have to wonder if people like Braitenberg really intended their > works to be perceived as a Disneyesque literal interpretation of > human-like emotions applied to machines, rather than an attempt to > discover more about *human emotions* through the use of fictional > machine examples. -- Gordon
Technical Details
Message-ID:
<385d988b.0@news.victoria.tc.ca>Path:
arc1-ca.pugleaf.net!unknown!not-for-mailReferences:
<385D6A30.19FE0307@gmccomb.com>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: fd6dd,83e4d27f33f04137 X-Google-Attributes: gidfd6dd,public X-Google-Thread: 1045ed,22902696b7981b17,start X-Google-Attributes: gid1045ed,public X-Google-Thread: 105832,83e4d27f33f04137 X-Google-Attributes: gid105832,public X-Google-Thread: f4366,22902696b7981b17,start X-Google-Attributes: gidf4366,public X-Google-Thread: 10ca9b,83e4d27f33f04137 X-Google-Attributes: gid10ca9b,public From: uj797@victoria.tc.ca (Arthur T. Murray) Subject: Re: Emotion in Robots Date: 1999/12/20 Message-ID: <385d988b.0@news.victoria.tc.ca> X-Deja-AN: 562842626 References: <385D6A30.19FE0307@gmccomb.com> Organization: Victoria Telecommunity Network Newsgroups: comp.robotics.misc,comp.ai.alife,comp.ai.philosophy,bionet.neuroscience,alt.support.chronic-pain X-Complaints-To: newsabuse@supernews.com