Google is working on better chatbots
Chatbots have a critical design flaw in that they are still incapable of engaging in genuine conversations with a real human. That's the major hurdle in chatbot design, but Google showed a concept of what they'll be like in the future.
Google added:
Cue in Maena, Google's new chatbot model trained on "2.6 billion parameter end-to-end neural conversational model." Maena can conduct sensible conversations and respond with more accurate and specific answers than current bots.
Moreover, Google claims that Maena's performance is close to human-levels of conversation, using an evaluation metric they call "Sensibleness and Specificity Average" or SSA.
We might soon have bots where we can converse akin to how Iron Man talks to his bots (J.A.R.V.I.S. and F.R.I.D.A.Y.).
Google added:
| "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." |
Moreover, Google claims that Maena's performance is close to human-levels of conversation, using an evaluation metric they call "Sensibleness and Specificity Average" or SSA.
We might soon have bots where we can converse akin to how Iron Man talks to his bots (J.A.R.V.I.S. and F.R.I.D.A.Y.).
