In a recent TED talk Peter Diamandis and Chris Anderson, cofounders of the X Prize Foundation introduced a new idea for an X Prize. Their TED talk proposed a new prize for developments in artificial intelligence.
The object of this new competition would be creation of robot capable of giving a TED talk.
In a written statement, X Prize chairman and CEO Diamandis stated that the competition could spur innovation in biological research, education, health care and “fields we have not yet even imagined.”
TED (Technology, Entertainment, Design) is a Vancouver, BC – based speaking series started in 1984. The TED series evolved into an annual series beginning in 1990, where talks began to be held in Monterrey, California. Since then TED talks have grown to a global series of speaking programs.
Most X Prizes have specific rules and conditions but this artificial intelligence competition requires only that the winner’s invention be capable of taking the stage and delivering a speech at a TED conference. The winning machine must deliver a talk so compelling that it earns a standing ovation. Finally, the winning machine must do those things without any human interaction.
The Foundation is still working out contest details. They are seeking public input to help with this.
This new competition is a distant relative of the Turing Test. In 1950 computer scientist Alan Turing devised a test to determine if a computer is truly intelligent.
The Turing test involved a human sitting at a computer terminal typing out questions. The questions would go to a human and a computer to be answered. If the human conducted the test could not tell whether a person or a machine was delivering the answers, the machine passed the test.
Architect and graphic designer Richard Saul Wurman invented the TED speaking series.
Not everyone is excited about this new challenge. Harvard University computer scientist Ryan Adams said that the set-up would reveal little about artificial intelligence. True intelligence, he said, is about adapting, learning, making decisions under uncertainty and achieving objectives. “Giving a speech and then ‘answering questions’ doesn’t address any of these issues.”
Murray Shanahan, a roboticist at Imperial College London, suggested a better test. Having the robot go to a stranger’s home and make tea would test motor control, object recognition and problem solving – “the hallmarks of intelligence.”
A system that answers questions after the speech would be even more compelling, said Shanahan.
The Loebner Prize already exists to address that question and answer challenge by rewarding the winner of a Q&A challenge.
The Robotics Challenge, run by the U.S. Defense Advanced Research Projects Agency, offers a $2 million dollar prize to the team whose robot is able to complete a series of eight challenges, including unrolling a fire hose, driving a car, and clearing debris from a doorway.
Many tasks that appear simpler than public speaking are quite difficult for robots. Climbing a ladder is one example of such a task. The challenge proposed at Diamandis’ TED talk does not address those artificial intelligence challenges with the new X Prize idea.
Diamandis admitted that the experts may be right. He said they haven’t established the rules and just wanted to put the idea out there for discussion.
The X Prize Foundation was established to encourage breakthroughs in science and technology. Their most famous challenge thus far offered $10 million to the first private organization that could fly a craft into near-orbit twice within two weeks.
According to a statement submitted to NBC News the prize amount for the AI challenge has not been decided but the winner would be judged on audience applause.
Diamandis and Anderson proposed a way to push the development of artificial intelligence with a huge prize. But, details of what the challenge would entail need to be decided.
By Chester Davis