Robots Teach Each Other New Tricks (technologyreview.com)

schwit1 writes with this story from the MIT Technology Review about a robot at Brown University who was taught to perform a task from another robot at Cornell. According to the article: "the ability to acquire and then share knowledge is a central component of human culture and civilization. A small milestone in the exchange of robot knowledge has now been demonstrated by two bots working in different academic research labs. Researchers at Cornell University previously devised an online game, called TellMeDave, through which volunteers can help train a robot to perform a task and associate different actions with commands given in everyday language. By guiding the robot through a task, a volunteer trains a machine-learning algorithm so the robot can perform the task again. And this learned behavior is stored in a central repository called RoboBrain that's accessible by other robots (see 'The World's First Knowledge Engine for Robots')."

Comments

That's not what the article says.

[Sowelu]

Robots, or rather a human-driven learning program, can contribute facts to a shared database. Any robot based on this system can pull facts out of that database. Having "one robot teach another" seems like it implies that they don't share a database, and instead communicate information through some other way. This article is pretty neat because of its fact database, but not as cool as gorillas teaching each other sign language.

Re:That's not what the article says.

[Nahor]

Indeed, schwit1 seems that have missed the point article, or at least didn't really explain clearly. The point of the is not about a robot teaching a other robot. That's just a save/restore.
And as you point out, it's not about the teaching part either, i.e. not about how the information is transferred from one robot to another, that's just simple shared database and a network connection.

It's about what is transferred from any robot to any other robot, and in particular, robots of different types, with different sensors and different actuators, so that the later robot can do the same job as the former.
In other words, the problem being solved is how to store the knowledge of one kind of robot so that it can be understood by another kind.