Moura and Kottur paper on non-human robot languages cited in The Atlantic


July 5, 2017

Recently, researchers at the Facebook Artificial Intelligence Lab found that their robots, which were being trained to negotiate with one another, started communicating in a non-human language when left to themselves. Other researchers including ECE Ph.D. student Satwik Kottur and ECE Professor José M. F. Moura have also observed machines formulate their own languages. In a 2017 research paper, Kottur, Moura,and their colleagues from the Georgia Institute of Technology and Virginia Tech describe their experience with two robots that eventually formed their own language by assigning values to colors and shapes. This research is important because it sheds light on how robots could communicate with each other in the future and how communication potentially emerged among humans in the first place. Kottur and Moura’s research was cited in an article for The Atlantic.

Related People:

José Moura