Ever since Facebook began its trial run for “M” in 2015, we’ve been anticipating the official roll-out of Messenger’s AI backed assistant. As of last Thursday the wait is over.
After last March’s updates that introduced a redesigned 3-level Persistent Menu and the ability to turn off freeform input, early this month Messenger launched its assistant “M Suggestions” for all iOS and Android users in the US.
Privately weaving itself into your conversation, M makes informed suggestions of Messenger’s internal features based on your conversational content, then displays these recommendations above the redesigned one-line message composer. For instance, if “Pay $16” is mentioned in a conversation, “M” can suggest the use of Facebook’s payment feature; when users talk about needing a ride somewhere, “M” will automatically display the “Get Ride” suggestion to access the Lyft and Uber-hailing options. Moreover, the artificial intelligence assistant can recognize relevant emotions through the typed sentences and suggest related Stickers.
“M Suggestions” can learn from users’ behaviors. If you frequently use its suggestions, “M” will keep pushing them on a regular basis; if you only use specific features such as “Start Plan” or Stickers, “M” can understand that only these recommendations are relevant to you and will only suggest these ones.
Can “M” invite chatbots into the conversation? Rumors abound regarding the likely imminent F8 announcement of integrating chatbots into group conversations. Why couldn’t “M” suggest adding a travel assistant chatbot to a conversation about your exciting travel plans with your friends in Bali next month?
Meanwhile, Facebook is keeping in action the original version of “M”, which has been tested by a small community of users since 2015. This older version of “M” will continue to build its knowledge of human interactions by answering questions based on what it has previously learned. We can expect “M”, the Facebook artificial intelligence assistant, to become more and more intelligent, and offer increasingly accurate and relevant solutions.
These articles can interest you