Researchers noticed a “divergence from human language” during chatbot interactions
June 15, 2017
Chatbots created at the Facebook Artificial Intelligence Research lab developed their own language without being directed by researchers.
As the bots’ development progressed, researched say they suddenly noticed a “divergence from human language” during interactions, forcing them to alter their model.
“In other words, the model that allowed two bots to have a conversation—and use machine learning to constantly iterate strategies for that conversation along the way—led to those bots communicating in their own non-human language,” notes The Atlantic’s Adrienne LaFrance.
A similar scenario was seen last year when the “Google Translate” service invented its own language to translate with.
“The online translation tool recently started using a neural network to translate between some of its most popular languages – and the system is now so clever it can do this for language pairs on which it has not been explicitly trained,” New Scientist’s Sam Wong wrotee. “To do this, it seems to have created its own artificial language.”
Most bots prior to Wednesday’s announcement were largely limited in their ability to carry out complex conversations, researchers say. Now, Facebook’s chatbots are not only capable of estimating and negotiating over an item’s “value,” but can even use deception in order to broker a deal.
Researchers said the bots “initially feigned interest in a valueless item, only to later ‘compromise’ by conceding it — an effective negotiating tactic that people use regularly” – a method developed by the bots on their own.
This article was posted: Thursday, June 15, 2017 at 5:05 pm