Seems like AI-based virtual assistants are the new happening things in the world of technology. After Bixby and Arbo that made some headlines in the recent times, Facebook Messenger is all set to welcome back M – its own AI-based virtual assistant.
Facebook’s first stab at an automated assistant, M could hold a coherent conversation, understand instructions and do your bidding. The early beta service, available to a handful of testers using Messenger, was powered by AI technology with a lot of help from invisible humans.
On Thursday, 6 th April 2017, almost two years after announcing the experimental feature, Facebook released a version of M to the general public. This M is not like the old M. Prepare to be underwhelmed.
Instead of a full-featured assistant, new M eavesdrops on your regular Messenger conversations and then pops in “suggestions”. It will work for iOS and Android users in the U.S. first, then come to other countries.
Facebook said that its messaging app, Messenger, will soon start showing users suggestions from a virtual assistant to Android and iOS users in the US. The artificial intelligence-powered assistant, M will suggest things it deems might be helpful to users based on the conversations they are having. This can include sending stickers, sharing your location with a friend to meet up, hailing a ride or sending money to friends.
Is your conversation too dependent on old fashioned words? M has some stickers to suggest. If you’re arguing with your pals about who the best Batman was, M can turn it into a poll. Some of the features are useful – if pretty standard in messaging apps. You can share your location mid-conversation or pay a friend back for last night’s margaritas. It also detects when you’re talking about heading out and offers a button to call an Uber or Lyft.
It’s a disappointing sequel to the early M experiment, but not surprising. Using real human contractors to call businesses, schedule appointments and answer complex questions is not cheap. Especially when you have 1 billion users. Current AI technology is just not as advanced, yet.
If you find M’s constant intrusions annoying, you can turn it off. Facebook will add more features over time, and it says M will learn about you the more you use it.
The more someone uses M, the “smarter” it gets – and if the suggestions are routinely ignored, it will stop providing them. While having a virtual assistant monitor your private messages can sound creepy, Facebook stresses that there is no advertising component to the move.
Facebook unveiled M in 2015, but until now it’s only been testing out the suggestions with a small percentage of Messenger users. Facebook is also launching a resource to help you spot false news and misleading information that spreads on its service.
The resource, similar to previous efforts around privacy and security, is basically a notification that pops up for a few days. Clicking on it takes you to tips and other information on how to spot false news and what to do about it.
Tips to spot false news include looking closely at website addresses to see if they are trying to spoof real news sites, and checking websites’ “about” sections for more information. Some sites might look like real news at first glance, but their “about” sections inform the visitor that they are in fact satire.
Adam Mosseri, vice president of News Feed at Facebook, said he hopes people will become “more discerning consumers” of news. The new feature is part of a broader plan by Facebook to clamp down false news stories, which gained outsized attention in the months leading up to the 2016 US presidential election.