The Future of Artificial Intelligence will not simply be limited to advanced numerical analyses; AI will also change the way machines ‘empathize’ and communicate with their human counterparts. This opens up a new frontier of opportunity for the service industry, and for the philosophy of robotics, in general.
So, talking robots, you say? That sounds a little too wacky. But then again, they said the same thing about the TV, didn’t they? To quote Darryl Zanuck during his 1946 outburst: ‘People will soon get tired of staring at a wooden box every night’…. I think it’s safe to assume that ‘they’ didn’t get tired.
But, having ‘conversations’ with machines is actually more common than you’d think: currently, 30% of all human interactions with technology have been through ‘talking’ with smart machines, with this type of technology already being able to respond successfully to around 80% of all customer queries. We know this phenomenon specifically through talking to ‘Chatbots.’ Traditionally, this type of ‘robot’ was designed to simply recognize the words within a customer’s typed-out message, and respond by following the linear logic of a preinstalled decision tree. But soon, chat bots will be programmed to ‘learn’ after every interaction, (a distinctly mammalian characteristic) and evolve strategies of customer interaction through forming their own logic. This inches into ‘Sci-Fi’ territory. Why? Because it’s the first step from transitioning AI from ‘Machines’ into ‘thinking’ beings.
Let’s take a step back from ‘Sci-Fi’ for a second and talk about a level of communication that we’re perhaps a little more familiar with: human interaction. Currently, it’s still more common to talk to a technical expert on the other side of the phone, than it is to be talking to a robot. Unfortunately, call centers can wrack up thousands of hours of call recordings every day. It’s a messy business, with employees tasked to manually mull over these recordings, looking mainly for two things: What are customers complaining about? And what is the refund rate for their products? In this case, AI involved in Text Analytics are able to scan transcripts of customer call data, and use what is known as ‘Sentiment Analysis.’ Programmed with the distinct purpose of understanding ‘How do they feel?’ processing customer interactions can gain insight into the ‘Sentiment’ behind the voice or text: if the customer is relating anger, contentment, or delight within certain products or services offered.
Indeed, the basis of an enhanced AI-based service industry relies on a robot’s ability to ‘empathize’ with the needs of their human customers. This is where the ‘Intelligence’ part of ‘Artificial Intelligence’ comes into play. There is a little-known sub genre of robotics which features ‘Natural Language Processing,’ also known as NLP, which gives a machine the ability to analyze and represent human language with the similar richness as a human would express him/herself. Before this, AI language interpretation was programmed to simply associate a symbol with a concept: specifically, words being represented as symbols. This was a good start, but in reality, language hardly functions on this ‘Symbol = Concept’ basis. This is because ‘words’ inherently depend on the context it is given. Through applying NLP, enhanced AI communications will not only understand the words themselves, but also the intent and context behind the words.
Another characteristic that makes AI a game changer within the service industry (other than sticking a C3PO behind every cash register) is its facility as a ‘Predictive’ field service. This gives machines ‘nonlinear’ thinking patterns, much like humans, and lets it anticipate incoming necessities and solve problems before they arise. Some examples include technologies which automatically rearrange employee schedules if, in the middle of the day a better way to deliver services arise. This type of adaptive thinking can help minimize the risk when unforeseen dangers arise, such as in the event of a gas leakage, where the AI can come up with a plan to minimize harm in the most cost effective way possible.
Nevertheless, SpaceX still hasn’t come out with a mass produced ‘R2D2’ model, just yet. But these small ubiquitous changes in Robot intelligence might just be the first signs of a brave new world, where human-like AIs are responsible for everything from delivering our pizzas to running our banks.