Facebook has recently launched its open-source and the largest ever chatbot Blender. According to few studies, it has a very high engagement rate and feels more humanistic.
The Research
Facebook has been researching from years in conversational AI. This is a blend of various set of conversational skills which includes:
Empathy,
Knowledge and
Personality together in a system.
The Achievement
Facebook Team published in a newsletter that they achieved this milestone through a new chatbot recipe that includes improved decoding techniques, novel blending of skills, and a model with 9.4 billion parameters, which is 3.6x more than the largest existing system. They also added that they are releasing the complete model, code, and evaluation set-up, so that other AI researchers will be able to reproduce this work and continue to advance conversational AI research.
Why Facebook made it Open Source?
Conversation is a talk, especially an informal one, between two or more people, in which news and ideas are exchanged. From decades, the AI engineers are working on building a system that can converse finely as Humans can with ease and grace showing sentiments and feelings. For this to happen, AI must be able to understand the broader context of the conversation. And for that, the system must be continuously upgraded which can be done easily when it will be open source.
Integrated Conversation Skills
Facebook has blended this chatbot with multiple conversational skills as we've discussed above. Now I will briefly discuss them:
This chatbot uses your personality to converse with you and make the conversation more engaging.
This chatbot uses a wizard of Wikipedia with the help of which it generated its knowledge base to answer your questions fluently.
It displays emotions, feelings and sentiments to make the conversation feel more humanistic by using empathetic dialogues.
And all these abilities and skills are blended so well that no one can discriminate from a real human.
The End Notes
The progress is exciting however they are far from achieving human level intelligence in dialogue system. But a open-source system will improve it much faster. I hope that the AI research community can build on this work and collectively push conversational AI forward. That's all in this article. Will meet you soon with another one.
Cheers!
Comments