Videos of student robot projects from U of Alberta's PSYCO 403X1 Research in Cognitive Science. Here's an answer to the burning question, "Do robots prefer Pepsi over Coke?"
Omni Brain
An exploration of the serious/fun/ridiculous - past/present/future of the brain and the science that loves it.
The Homunculus
The Omnibrain is a psychology graduate student at an online university. He hopes that the three weeks and $29.95 that he is spending on his Ph.D. will get him a job at a Tier 1 research university. Do online universities have postdocs? Ok...just kidding, The Omnibrain is a real graduate student at a real school somewhere in the continental United States - or maybe Europe.
Search
Recent Posts
- A case where PETA might actually be helpful
- Get hypnotherapy with a cat
- Optical Illusion Video - Different Colors?
- Someone stole the Nobel Prize!
- A great moving depth illusion
- A scientific study of Superman's flight
- How Supermans X-Ray vision works
- Lab tech afraid Ted Williams' cryogenically frozen head about to become zombie beats it to death
- The Anatomy of Japanese Mega Monsters
- It's a dog eat dog world
Recent Comments
- analis on Smokin' Weed. Is it good or bad for you?!
- louis vuitton girl on The Brain That Wouldn't Die in the public domain
- DDos Protection on Galileo flipping off the Catholic Church from beyond the grave
- DDos Protection on How moving your eyes in a specific way can help you solve a complex problem without even realizing it
- DDos Protection on Best Visual Illusion of the Year 2007
- DDos Protection on Psychopharmacologist Song
- DDos Protection on How wearing underwear led to increased medieval literacy
- DDos Protection on The Purple Nurple Optical Illusion
- sean on Transcranial Magnetic Stimulation At Home.
- sescotampan on Is a flying carpet possible?
Archives
- October 2009
- September 2009
- February 2008
- January 2008
- December 2007
- November 2007
- October 2007
- September 2007
- August 2007
- July 2007
- June 2007
- May 2007
- April 2007
- March 2007
- February 2007
- January 2007
- December 2006
Blogroll
- A Blog Around the Clock
- The Anterior Commissure
- Cerebrum: The Dana Forum on Brain Science
- Channel N
- Cognitive Daily
- Cognitition and Language Lab
- Developing Intelligence
- The Frontal Cortex
- Improbable Research
- Milligram
- Mind Hacks
- Mixing Memory
- The Neurocritic
- Neurophilosophy
- Phineas Gage Fan Club
- PsyBlog
- Pure Pedantry
- Retrospectacle
- Sequitur
- Shrink Rap
- Tierney Lab
- World of Psychology
Glial Cells
Axons
« Time Graph | Main | Do you have a war on drugs problem? The drug Incarcerex may be for you. »
Multimedia Friday 06/29/2007
Category: AI • Academia • Neurotech & Robotics • Technology • Video
Posted on: June 29, 2007 8:30 AM, by Sandra Kiume
Share this: Facebook Twitter Stumbleupon Reddit Email + More
TrackBacks
TrackBack URL for this entry: http://scienceblogs.com/mt/pings/44247
Comments
Wow, after watching that video one cannot help but think that intelligent robots are not that far away...
Posted by: The Neurocritic | June 29, 2007 7:31 PM
The link doesn't give us any information about the experiment. Was this done by AI or by human control? If the former, wow...it's not too hard to replace "Coke" with "human" and the claw with a laser.
Posted by: Brandon | June 30, 2007 3:43 AM
Hi Brandon - this was done by autonomous robots, though calling it AI is premature. It actually wouldn't be that easy to program it to "kill all humans" because how do you define a human? It's a somewhat abstract idea that we haven't even figured out ourselves, and robots need very specific instructions and parameters. Other robots can be shaped like humans, other primates can radiate body heat, humans come in many sizes and can shield themselves, etc. so mistakes could be made. Object recognition is improving but we're not quite at that stage yet.
We could give them lasers, though. :)
Posted by: Sandra | June 30, 2007 8:51 PM
Looks like they're using a Lego RCX brick, with light/colour sensors used to differentiate between cans, and a sensor pointed at the floor to stay within the taped off boundary. The 'AI' would be something like...
1 - Look for object in proximity, and charge at it.
2 - If the object doesn't reflect light like a coke can, then back off, turn around, and go to 1.
3 - If floor reflectance drops, turn around, and go back to 1.
It's still really quite fun to make robots like this! However, the object recognition isn't very robust, so I'd expect things to go wonky if the lighting conditions changed.
Sandra's comment does remind me of the autonomous rescue robots in the Robocup competition - you would just have to change rescue to kill....
Posted by: David | June 30, 2007 11:24 PM
Wow Sandra, it's very disturbing that you've put that much thought into this. Even so, excellent work.
Posted by: Brandon | July 1, 2007 2:04 AM
Heh. I recently saw some combots in death matches at RoboGames and the thought occurred to me. But I can't take credit for all that, I read an interview with Paul Almond in Machines Like Us in which he talks about Asimov's famous three laws of robotics (can't harm a human, etc.) and how difficult that could be to teach a system. I'm sure many researchers and philosophers have considered the question.
Check out another ScienceBlog, Developing Intelligence - Chris probably knows more.
Posted by: Sandra | July 1, 2007 3:47 AM