Science
Related: About this forumRobot learns to recognise itself in mirror
A robot named Nico could soon pass a landmark test - recognising itself in a mirror.
Such self-awareness would represent a step towards the ultimate goal of thinking robots.
Nico, developed by computer scientists at Yale University, will take the test in the coming months.
The ultimate aim is for Nico to use a mirror to interpret objects around it, in the same way as humans use a rear-view mirror to look for cars.
http://www.bbc.com/news/technology-19354994
The next test will be to check if his fly is open.
Ezlivin
(8,153 posts)Now let's see how he does on the rest.
applegrove
(118,837 posts)Jim__
(14,088 posts)But the article states:
So far the robot has been programmed to recognise a reflection of its arm, but ultimately Mr Hart wants it to pass the "full mirror test".
The so-called mirror test was originally developed in 1970 and has become the classic test of self-awareness.
I've tried to find some more information on the "full mirror test," but everything I find is about this robot. My problem is that I'm not convinced that a robot acting as if it recognizes itself in a mirror is a test for actual self-awareness.
I agree (i)t is a spatial reasoning task. But, say the robot is named RobotI. If the robot is taught about the reflective properties of mirrors, and given parameters that allow it to identify RobotI in the mirror and how to identify the location of items in the room by looking at the reflection in the mirror, I would think that the robot could pass this test without actually being self-aware.
Again, I think it's a tremendous accomplishment to get a robot to recognize itself in a mirror; I'm just not convinced, based on what I've been able to read, that it is a real test for self-awareness. But, it may be that what they're calling a test of self-awareness means something different from what I think.
Posteritatis
(18,807 posts)It's still neat though, I think only a half dozen or so species other than the great apes have passed it so far. (Weirdly enough, the jury's still out on one of them - pigeons, of all things.) Humans don't until they're around two.
And of course determining self-awareness for artificial intelligences is going to be a whole other story in general, if certainly an incredibly interesting one.
Harry Monroe
(2,935 posts)I've seen the "Terminator" series. This ain't gonna end well.