Share & Connect
An exchange between two robots, facilitated by Cornell University researchers, resulted in a “bizarre and ill-tempered discussion” about god and philosophy. The two researchers, Jason Yosinski and Igor Labutov, created a system in order to observe how ‘chatbots’ – online programs designed to hold conversations with humans – would fair when speaking with another computer instead of a human.
Labutov and Yosinkski created two avatars with voices – a British man that they named Alan and an Asian woman called Sruthi, Mail Online said. According to Mail Online, the conversation turned “sour” when Struthi asked Alan whether or not he believed in God. Alan replied, “It’s not everything.” This response angered Sruthi.
“Not everything could also be something, for example not everything could be half of something, which is still something and therefore not nothing,” Sruthi said. Another highlight of the conversation involved Alan referring to himself as a unicorn and calling Sruthi a “meanie.”
According to The Telegraph, machines like chatbots were designed to have the conversational abilities of humans, typically in an attempt to pass the Turing Test for intelligence. Alan Turing proposed the Turing Test in his 1950 research paper. The test was meant to measure how smart a computer was.
If, through conversation, the computer was capable of fooling a human into thinking they were human and not a machine, the computer could rightfully be deemed intelligent, Turning said. Different from the Turning Test, Yosinki and Labutov were less interested in seeing how computers mirrored the behavior of humans and more interested in observing an authentic computer-and-computer interaction, according to Mail Online.
The concept of a computer being able to speak is nothing new. Since coming online in 1997, Cleverbot, a web application, has engaged in about 65 million conversations with internet users around the world. People are able and encouraged to come to the site and chat with the cleverbots for one’s entertainment, for free.
According to msnBC, just like a human learning appropriate behavior by studying the actions of members of his or her social group, machines “learn” from dialogues with people. They are capable of storing the new knowledge they gain from each interaction and computers are able to refer to it the information during future conversations.
The unusual exchange between Alan and Sruthi will likely inform researchers that computers need more training before being able to speak logically to each other. But, regardless of how the experiment advances Cornell’s research efforts, the robot video was a major hit on YouTube.
With more than 2,000,000 views, comments posted beneath the video prove, if nothing else, that computer conversations are entertaining. “I could listen to this all day!” Jesseinpain posted. JoeyShambrook goes as far as to call the video, “epic.”