A crowdsourced artificial chat partner that’s smarter than Siri-style personal assistants
September 10, 2012
The Chorus system. User requests are forwarded to crowd workers, who then submit and vote on responses. Once sufficient agreement is reached, responses are made visible to users. The crowd’s working memory is updated by workers selecting lines from the conversation or summarizing important facts. (Credit: W.S. Lasecki, Rachel Wesley, Anand Kulkarni, J.P. Bigham)
Personal assistants such as Apple’s Siri may be useful, but they are still far from matching the smarts and conversational skills of a real person.
Researchers at the University of Rochester have demonstrated a new, potentially better approach that creates a smart artificial chat partner from fleeting contributions from many crowdsourced workers, Technology Review reports.
When people talk to the new crowd-powered chat system, called Chorus, using an instant messaging window, they get an experience practically indistinguishable from chatting with a single real person.
Yet behind the scenes, each response is the result of tens of people paid a few cents to perform small tasks: including suggesting possible replies and voting for the best suggestions submitted by other workers.
Tests where Chorus was asked for travel advice showed that it could be smarter than any one individual in the crowd, because around seven people were contributing to its responses at any one time.
Comments (5)
by donjoe
Resistance is futile.
by grettir76
Digital Self on the way?
by Starheart
As far as I am concerned, while the concept may sound silly, if the system actually adapts its internal databases with crowdsourced feedback in mind to the point where eventually it will pick relevant responses without human intervention, then this is a prime example of something not yet attempted in AI development: learning by imitation. Which is how just about every single child acquires basic skills.
by Marcos Marin
not only, even monkeys do it.. anything that isn’t simply reacting instinctively actually! This should tell you something about the hubris and quality (if only the former was justified by the latter) of decades of AI ‘researchers’. In fact, the least verisimilar aspect of the Prototype trailer (also on today’s issue of the newsletter) was the guy was from MIT, which can’t be trusted to produce anything much higher than a retarded cockroach (to use Michio Kaku’s term, lol)
by GAUSS
Quite right. Very interesting stuff. It’ll be cool to see when systems like these sort of spread their wings and become more autonomous.