New Delhi: Google has announced its new chatbot ‘Meena’, which the company claims that it is the best chatbot anyone might have ever interacted with till date.
The company says that the bot’s neural network has about 2.6 billion parameters and so Meena will be capable of engaging in a conversation about any topic on the Earth. Google scientists in a paper recently showed that Meena can conduct conversations that are more sensible and specific than existing state-of-the-art chatbots.
To put Meena, an end-to-end trained neural conversational model, to test, the scientists used a new human evaluation metric for open-domain chatbots, called Sensibleness and Specificity Average (SSA), which captures basic, but important attributes for human conversation.
The Google Chatbot uses a seq2seq model along with a variation of the famous Transformer architecture and is trained on 40 billion words. Google has claimed that Meena is better than Siri, Alexa, and other voice assistants.
However, in a blog post, Google confirmed that the chatbot has not yet achieved perfection.
“While we have focused solely on sensibleness and specificity in this work, other attributes such as personality and factuality are also worth considering in subsequent works. Also, tackling safety and bias in the models is a key focus area for us, and given the challenges related to this, we are not currently releasing an external research demo,” said researchers.
What is a chatbot?
A chatbot is a machine learning language or software that uses AI, letting textual conversation with humans, via a live chat interface.