Personal assistant that needed human help gets the boot
A spokesperson for the social network told The Register on Monday: “Today we shared with the people who have access to our M closed beta project, first announced in August 2015, that January 19th is the last day the service will be available.”
It’s no surprise that the project failed, really. There were warning sirens going off in April last year when it emerged that when M faced questions it couldn’t understand, it would fall back to a team of humans to answer the queries. There was obviously no way it was going to scale out from its limited beta test of about 2,000 people in the US to the billion or so people who log into Mark Zuckerberg’s empire every day.
M was heralded as a Siri-like personal assistant that you could natter to via text message, and that it would be useful and smarter than today’s clumsy rule-based chatbots by answering questions and carrying out requests for stuff.
Over time, however, it was relegated to a research experiment when it was clear M was not much better than other dumb bots. Some of the technology will continue in the form of M suggestions, which pop up during Messenger chat to suggest stickers, booking a cab ride, and so on.
Facebook didn’t want to go into why M is being retired. The Silicon Valley biz has only told us that it launched the project to “learn what people needed and expected of an assistant, and [it] learned a lot.”
A spokesperson declined to explain specifically why M was being killed off, how many human handlers it had, and how much money was blown on research and development for the project. “Regarding your questions, I don’t think we have anything else to add,” the rep told El Reg.
In an earlier statement to the press, first shared with The Verge, Facebook said: “We’re taking these useful insights to power other AI projects at Facebook. We continue to be very pleased with the performance of M suggestions in Messenger, powered by our learnings from this experiment.”
The lesson that should be learned from M is clear: it’s 2018, and digital assistants still suck. ®