Servo Magazine ( November 2016 )
Facebook and Microsoft Messenger Bots: It’s Déjà vu All Over Again
2016 is the year of the chat bot — a forecast supported in part by the launch of messenger bots for Facebook and Microsoft platforms, among others. Chat bots — introduced nearly 50 years ago — are software agents that can create text dialogue that resembles something that a human might produce.
Eliza — a bot programmed to mimic a psychotherapist — was one of the earliest examples of natural language processing (NLP) that made it to the masses. There are numerous web based versions available if you’d like to try your hand at self-psychoanalysis.
The last big push for NLP based bots was two decades ago, when web based software bots were seen by many as the marketing, sales, and customer service forces of the future. Why pay for call centers to handle complaints when a chatter bot can do the same for pennies on the dollar? Banks began the practice of handling customers with large balances with human operators and everyone else with chatter bots.
This made sense in a perverse way. Worst case scenario, a bank customer with a small balance would become upset and move their savings to another bank. This left the high value customers, which was better for the bank’s bottom line.
When I was involved with a startup that leveraged NLP technology and medical chatter bots in the ‘90s, the focus was on personal avatars, speech recognition and generation, and the end of email communications as we knew it.
Well, post-bubble, that company (and a lot more like it) failed to survive, and the excitement about chatter bots calmed to a murmur.
The latest announcements regarding chatter bots once again taking over the world of e-commerce may be more than simply Déjà vu. Two decades ago, we weren’t all toting smartphones, using them to pay for items at the point of sale, and ordering from Amazon and other online retailers on our commute home.
Today, there just may be enough of a user base accustomed to texting everyone and every business to make e-commerce using chatter bots viable.
I’m, of course, curious to see to what extent the various tool boxes announced by Microsoft and others can be applied to mechanical robots as well. Good speaking manners would be a big plus in creating an emotionally intelligent interface to a social robot.
Similarly, I expect the latest generation of chatter bots — tied to voice recognition and generation — to make a significant impact in the car market.
After all, what good is a self-driving car if you can’t have an intelligent conversation with the driver? SV