[ad_1]
Facebook has launched a new chatbot that it claims can demonstrate empathy, knowledge and personality.
“Blender” was trained using available public domain conversations that included 1.5 billion examples of human exchanges.
The social media giant said 49% of people preferred interactions with the chatbot, compared to another human.
But experts say that training artificial intelligence (AI) using a platform like Reddit has its drawbacks.
Numerous problems arose during the longer conversations. Blender sometimes responded with offensive language, and other times made up facts entirely.
The researchers said they expected other models to address some of these problems.
“We believe that launching models is essential to allow a complete and reliable view of its capabilities,” said a Facebook spokesperson.
- A chatbot took me out of a “really dark place”
- Can computer translators beat a human?
Artificial Intelligence expert Dave Coplin, executive director of consulting firm The Envisioners, said Blender was a “step in the right direction,” but pointed to two fundamental issues that still need to be overcome.
“The first is how complex it is to replicate every nuance of a human attribute, such as the ability to carry on a conversation, a skill that most three-year-olds can master,” he told the BBC.
“The second is about the relationship with the data used to train the model and the results generated by the model.
“As great as Reddit is, training algorithms based on the conversations you find there will give you a lot of straw in the wheat.”
Facebook also compared Blender’s performance to the latest version of Google’s chatbot, Meena.
He showed people two conversations, one with Blender and the other with Meena.
The conversations included a wide range of topics, including movies, music, and veganism.
Facebook said 67% of respondents thought that Blender sounded more human than Meena.
“We achieved this milestone through a new chatbot recipe that includes improved decoding techniques, a novel combination of skills, and a model with 9.4 billion parameters, which is 3.6 times more than the largest existing system.
“This is the first chatbot that combines a diverse set of conversation skills in one system.”
“Building a truly intelligent dialogue agent who can chat like a human remains one of the biggest open challenges in AI today.”