Google is developing a bot that can hold almost human degree conversation



	February 3, 2020

With more than 300,000 active chatbots on Facebook Messenger, Facebook hope that they might gain traction.

But unfortunately, they didn’t. Part of the problem is that people don’t welcome talking to a bot. another common problem is that bots in general aren’t very engaging.

Take for example your home speaker device, unless you say the exact right thing, the device might not respond or understand.

That is what every bot and AI developer is trying to solve. And this week, Google has shown a glimpse of their new bot system. The system will recognize more patterns and hold an almost human degree conversation.

Google officials stated:

"Current [...] chatbots have a critical flaw - they often don’t make sense. They sometimes say things that are inconsistent with what has been said so far, or lack common sense and basic knowledge about the world. Moreover, chatbots often give responses that are not specific to the current context."

Google new bot system is called Meena, and is trained on “2.6 billion parameter end-to-end trained neural conversational model", and can conduct more sensible conversation than existing bots.

Google says the Meena's performance is actually closing in on human conversation - using an evaluation metric that it calls the "Sensibleness and Specificity Average (SSA)", which captures "basic, but important attributes for human conversation".

That could be a significant evolution for bot usage. If bots can hold more genuine conversations, that could make them more engaging, and improve adoption, either through novelty or practicality.