[Home]   [Full version]  

New Cornell institute will apply artificial intelligence to decision making and data searches

May 19 ,Technology


Suppose the computer from the starship Enterprise or the HAL 9000 from "2001, A Space Odyssey" had been scanning intelligence data four years ago. Perhaps it would have made the connection humans missed between terrorists and flight schools. Or suppose such a computer were designing airline flight schedules: You might get home for Christmas a little faster.

These are just some of the possibilities of "artificial intelligence," or AI, which is not really about making computers that talk back but rather about using computers for the things they are good at: dealing with massive amounts of data or problems with a vast number of choices. These are the sorts of problems that are being examined by Cornell University's new Intelligent Information Systems Institute, launched this year with a $5 million, five-year grant from the U.S. Air Force Office of Scientific Research (AFOSR). In keeping with university policy, none of the research will be classified.

Other problems on the table include game theory, information retrieval and automatic verification of software and hardware, according to Carla Gomes, Cornell associate professor of computing and information science and applied economics and management, and director of the new institute. Some 20 other Cornell faculty members are affiliated with the institute, Gomes said, including not only computer scientists but also faculty in operations research, applied economics, mathematics and engineering. Visiting scientists will be more the rule than the exception, she added, with at least 10 scheduled to arrive this summer.

The institute dedicated its new facility in Upson Hall April 21 with a ribbon-cutting attended by AFOSR officials and researchers from the Air Force Research Laboratory in Rome, N.Y., with which Cornell has close research ties. The laboratory also collaborates with Cornell in the Information Security Institute headed by Fred Schneider, Cornell professor of computer science.

The newly remodeled space includes computer terminals that provide access to the institute's dedicated computer cluster, a high-performance computer made up of a cluster of 12 Intel processors operating in parallel. But the main purpose of the spacious facility, Gomes said, is to inspire collaboration among scientists. "There are whiteboards everywhere," she pointed out, "and open workspaces, stimulating the exchange of ideas."

The creation of the institute has created a buzz in the academic community, according to Robert Constable, Cornell's dean of computing and information science. "At one time Cornell wasn't even present in AI, but now we're being recognized as one of the top places, and we're seeing it in the applications to graduate school." This year, he said, there have been more applications from students wanting to work in AI than any other field, and more of those accepted have decided to come to Cornell than ever before.

Source: Cornell University

Related stories:

Distinguishing between 2 birds of a feather
The bird enthusiast who chronicled the adventures of a flock of red-headed conures in his book "The Wild Parrots of Telegraph Hill" knows most of the parrots by name, yet most of us would be hard pressed to tell one bird from another. While it has been known for a long time that we can become acutely attuned to our day-to-day environment, the underlying neural mechanism has been less clear.
Researchers discover how antidepressants and cocaine interact with brain cell targets
In a first, scientists from Weill Cornell Medical College and Columbia University Medical Center have described the specifics of how brain cells process antidepressant drugs, cocaine and amphetamines. These novel findings could prove useful in the development of more targeted medication therapies for a host of psychiatric diseases, most notably in the area of addiction.
Roadrunner supercomputer puts research at a new scale
Less than a week after Los Alamos National Laboratory's Roadrunner supercomputer began operating at world-record petaflop/s data-processing speeds, Los Alamos researchers are already using the computer to mimic extremely complex neurological processes.
'Nanoglassblowing' Seen as Boon to Study of Individual Molecules
While the results may not rival the artistry of glassblowers in Europe and Latin America, researchers at the National Institute of Standards and Technology and Cornell University have found beauty in a new fabrication technique called “nanoglassblowing” that creates nanoscale (billionth of a meter) fluidic devices used to isolate and study single molecules in solution—including individual DNA strands. The novel method is described in a paper posted online this week in the journal Nanotechnology.
What horses can tell us now about the coming human flu pandemic
Stored safely in a freezer at Cornell's James A. Baker Institute for Animal Health are samples of the virus thought to be most like the one public health experts expect someday to afflict record numbers of the world's population. The virus was collected in 1973 during an outbreak of equine influenza at a Florida racetrack. Dorothy Holmes, an infectious disease specialist in Cornell's College of Veterinary Medicine, had obtained samples of the virus with the intention of using it to create nasal spray vaccines for horses.
Magnetic levitation gives computer users sense of touch
Computers, long used as tools to design and manipulate three-dimensional objects, may soon provide people with a way to sense the texture of those objects or feel how they fit together, thanks to a haptic, or touch-based, interface developed at Carnegie Mellon University.
Migration from Africa left mark on European genetic diversity
Human migration from Africa to Europe more than 30,000 years ago appears to have left a mark on the genes of Europeans today.
Predicting the perfect predator
Garlic mustard has become an invasive species in temperate forests across the United States, choking out native plants on forest floors and threatening ecosystem diversity. University of Illinois ecologist Adam Davis has created a computer model that in combination with quarantined research tests he believes will be able to predict the perfect predator -- a pest that can be introduced into a forested area that will help reduce the garlic mustard population.

News discussion:

Technology news

[Home]   [Full version]