Science

Engagement in Humans and Robots

Posted 25 Jul 2005 at 15:38 UTC by steve Share This

Engagement is a term coined in 1999 by Alan Bierman to describe the process by which indivudals start, maintain, and end a perceived connection to each other. Engagement involves both verbal communications, gesture recognition, face recognition, gaze tracking, and other phenomena. One problem with applying this concept to robotics is that no one has really studied engagement in humans enough to fully understand how it works. A new paper titled Exploration of Engagement for Humans and Robots (PDF format) takes a look at the details of engagement in both humans and robots. The research was done by MIT's Media Lab and the Mitsubishi Electric Research Lab.


Give Your Robot A Personality, posted 26 Jul 2005 at 06:10 UTC by AI4U » (Observer)

Engaging the Robot AI Mind in Win32Forth is deceptively simple, inasmuch as all communication between human and robot occurs initially via the keyboard. The Forthmind has concepts of self and of other, so that it perceives the input word of "you" as referring to the robot and when you speak of "I" it talks to you on-screen as "you."

If you ask the AI robot, "do you love me" (without punctuation), it only knows the answer if you have already told it the answer. Although an Emotion Module is in preparation that would let the robot feel love and hate and other emotions, right now the main effort here is in getting the robot to think logically and not make spurious associations.

Mind.Forth is a kind of Seed AI but it is not meant to be a Friendly AI as such, because there is no way to guarantee friendliness on the part of an autonomous, intelligent robot. Just as human parents could give birth to a murderous Hitler or GWBush or Stalin, likewise there is no guarantee that your mind-equipped robot might not try to take over the world. It is the duty of human society as a whole, not of individual AI developers or robotics hobbyists, to decide whether we humans will permit a parallel society of robots to emerge alongside our own society in Joint Stewardship of Earth.

Not a community member, posted 26 Jul 2005 at 10:48 UTC by The Swirling Brain » (Master)

This AI4U is not behaving like a community member. Therefore, I propose that this guy be removed from the robots.net. If you are one of the people who have certified him please remove your certification. Or, if you're one of the people who have certified the person who certified him, remove their certification also. Come on guys this has gone on long enough. It's your job to keep up with certifications and this is a perfect example to remove certification and keep robots.net a great place to visit. Also, it would be nice if this you would remove certifications from persons who haven't logged on in a very long time as it appears some of these haven't. I know it seems cruel, but he's giving everyone a headache and something needs to be done.

hmm play nicely..., posted 26 Jul 2005 at 11:17 UTC by dogsbody_d » (Master)

Come on SB, it does seem like we are slowly training him. Is it possible to have his posts moderated? This one was very nearly on topic, and while the meanderings at the end weren't, they were actually quite relevant to the "are machines evil" religion-type chats we've been having.

AI4U - I think you've maybe missed a trick with the engagement business. A strictly keyboard interface has a crucial chunk missing from this. It's impossible to distinguish between a gap in the conversation, and someone having left the keyboard for good.

Don't you think there's a problem trying to develop true AI that has no senses other than you as a voice in its head? Have you seen the Metallica video for "One?" You should consider writing a web module that would allow your "Mind" to find things out for itself, maybe even let it post comments on blogs...

barking up the wrong tree, posted 26 Jul 2005 at 11:49 UTC by c6jones720 » (Master)

Perhaps there should be more emphasis on using mixed teams to come up with solutions. It strikes me that this is more psychological, maybe we should get psychologists to solve this for us and then use their findings on our machines afterwards.

See more of the latest robot news!

Recent blogs

15 Nov 2014 mwaibel (Master)
14 Nov 2014 Sergey Popov (Apprentice)
14 Nov 2014 wedesoft (Master)
5 Nov 2014 shimniok (Journeyer)
5 Aug 2014 svo (Master)
20 Jul 2014 Flanneltron (Journeyer)
3 Jul 2014 jmhenry (Journeyer)
3 Jul 2014 steve (Master)
2 Jul 2014 Petar.Kormushev (Master)
10 Jun 2014 robotvibes (Master)
X
Share this page