Now playingSpaceCollective Where forward thinking terrestrials share ideas and information about the state of the species, their planet and the universe, living the lives of science fiction.Introduction Featuring Powers of Ten by Charles and Ray Eames, based on an idea by Kees Boeke.
"Native Americans, Aboriginal Australians, Taoists, Zen Buddhists, Tantric
meditators, and mystics everywhere do not think of the Dreaming world
as an 'un'-conscious. For these peoples, the sentient Dreaming world is
the basic reality. Though marginalized and invisible to mainstream cultures
today, Dreamtime has been the essential reality for people from the
beginning of time."
-Arnold Mindell
The setting sun glints off the Amazon River and numerous lakes in its floodplain in this astronaut photograph from August 19, 2008. About 150 kilometers of the Amazon is shown here, about 1,000 kilometers inland from the Atlantic Ocean.
This image was acquired on August 19, 2008 by the by the Expedition 17 crew of the International Space Station.
South of Khartoum, Sudan, where the White and Blue Nile Rivers join, a dizzying arrangement of irrigated fields stretches out across the state of El Gezira. The several bare-looking patches are small villages.
This image was captured by the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) on NASA's Terra satellite on December 25, 2006.
The Robotic Musicianship Group at Georgia Tech Center for Music Technology just blew our minds with some videos depicting robots playing music with real people. Great, you say. Some fake robot machine can, like, bang around on a drum or something. Not so fast, doubters.
These robots, developed with funding from the National Science Foundation, listen to humans creating music in real time and play along with them. One might say they improvise.
They can't pass the actual Turing test, in which a robot must fool a human into thinking it is also a human during a conversation.
But musical improvisation is another kind of a conversation and I, a human, would believe that the impromptu, non-predetermined parts these robots play were played by other humans. By that standard, the Georgia Tech team's robots have already passed the musical Turing test (assuming that I'd feel the same way if I were playing along instead of watching the robots in a video, and I'm fairly certain I would).
We asked professor Gil Weinberg, head of the program, how these robots manage to parse what humans are playing, and how they manage to play along. How do they figure out which parts to play? As it turns out, the process is somewhat analogous to the way Deep Blue plays chess: by carefully examining its options and then evolving them like biological species to see which one best fits a changing musical environment.
"The processing allows [the robots] to analyze and improvise," said Weinberg via telephone. "In one of the applications, we use a genetic algorithm... You have a population of something, and then you do mutations to all of these little things — in my case it's musical motifs — mutations and cross-breeding between the musical genes, in our case, and then you have a new population that better fits to the environment.
He continued, "Very fast, it runs [about] 50 generations of mutations that are cross-bred between the genes and tests whether this is similar to a motif that the saxophone player played, for example. And it plays something back that is a combination of musical genes of what the saxophone player played, what the piano player played — something that is unique that only can be the product of genetic algorithm."
The results are fairly astounding. Haile, the drumming robot has been around for a couple of years, but Shimon, the marimba playing robot unveiled in early November, handles melody in addition to rhythm. One of the next steps, says Weinberg, is to give the robots to look at whichever human collaborator is playing the most interesting part.