A Beginners Guide to ChatBots: Are They Taking Over the World?

When you connect with a business nowadays, chances are your first point of contact is a bot. Whether it’s the traditional automated phone messages that ask you to tell them what you’re calling about or a seemingly sentient Chatbot that you’re texting with on Facebook messenger, there’s no doubt that the future of B2C seems to be with bots.

Since the integration of the “first” chatbots on AIM (American Online Instant Messenger) this technology has been on an upward path. Today, chatbots are used for all kinds of businesses from law firms, to food delivery, and even in banking. But not all bots are created equal.

Voice to Text Bots

Google Home and Alexa are becoming our one-stop-shop virtual assistants. These technologies are much more advanced than a predefined chatbot, as they use natural language understanding and speech recognition to properly function. However, this technology is not exclusive to these personal assistants. It can also be used to make a chatbot conversational and “smart”. To help with this, both Google and Amazon have released their speech recognition technologies to chatbot developers. This in turn allows Alexa and Google Home to improve different queries within multiple languages, societies, and cultures.

The majority of natural speech recognition is based on to-do lists, calendar scheduling, weather reports, and so forth. These are all predefined items that Alexa or Google understand. It’s simply the start for virtual assistance and home automation. Since these devices and operating systems don’t know everything (yet), they rely on extended third-party programs. Using programs, such as If This Then That (IFTTT), virtual assistants can easily automate repetitive tasks. Let’s say we want an automated text message when someone rings the doorbell; this can all be integrated through a virtual assistant and third-party applications.

But, how do these bots understand voice-based commands? Applications like Apple’s Siri will convert the speech into text, and revert the response of text into speech. The majority of these voice-to-text applications will use an Application Programming Interface (API), such as Google’s Speech API. The text-based or voice-to-text bots will know what API to call, information to fetch, and in which ways to respond. Retrieval-based bots are completely different, as they use preset responses; for example, these include bots that always start the conversation with “Hello, how may I assist you?”

Artificial Intelligence

The existing applications and virtual assistants are a long way from the high-tech supercomputer artificial intelligence (AI) of the movies. But, why do Alexa and Google Home appear so smart? Well, the majority of these virtual assistant technologies use various phrasings of a question to fetch an answer via the Internet. Alexa and Google Home may know a lot through fetched data, but they are only as good as the developers that made them and the data on the internet. The truest sense of AI is a machine understanding a question, and if not understood, clarifying the question to determine the answer. Machines classified as AI need to learn and understand new information, without human assistance. Nevertheless, AI isn’t taking over the world any time soon; it’s here to collaborate with people and enhance our experience. Something that’s programmed by humans, and fetches data that humans wrote online, isn’t too smart, yet.

As stated, the majority of AI today is action-based, which means it’s using natural language to process a command. This type of AI will learn by usage and user interaction. If the data is set to understand “Pay bill” for a credit card bot, then it will be able to understand “Pay full statement.” The credit card bot learns from numerous queries, and over time, understands natural language meanings. But, would a credit card bot take us over? Plainly no, these types of action-based bots still have parameters, and predefined data fields to fetch.

Text Based Bots & Messenger Programs

Text-based messaging bots also use natural language, and are commonly custom built with existing messenger applications. The text-based bots can be found on messenger programs, such as Slack and Facebook Messenger. Users may want to have a full conversation with a bot, when in reality, text-based bots are better-suited for individual tasks. Want the company to be notified when a fresh pot of coffee is ready? This is no problem for a custom built bot linked to your company slack channel. Bots are perfect for repetitive tasks.

Bots Are Taking Us Over!

They really are everywhere: in our automated banking, food delivery, weather feeds, online shopping carts, and more. The world is becoming more automated, and our human experience is being simplified through bot technology.

So, are bots truly taking over the world? Let’s think back to the 2004 movie, I, Robot when Del Spooner is speaking to Sonny the robot and says — “Human beings have dreams. Even dogs have dreams, but not you, you’re a machine. An imitation of life. Can a robot write a symphony? Can a robot turn a canvas into a beautiful masterpiece?” The answer to this, at least now, is no. But it does help us imagine what bots can become.

Bot technology will advance to the point where they can replicate our way of speaking, writing, and even thinking. It will advance to the point of deducting advanced questions, and producing an artificial answer. But, automated and bot technology won’t be able to dream, create, or relate on a real human level. As time goes on, we’ll depend on bots for efficient, non-emotional, and productive “in-the-box” styles. These bot styles will become commonplace in even more industries and as humans it will be our job to think outside-the-box, and keep bots from completely taking over our lives. We need to continue to dream, create, and relate on a human level, with everyone.

Do you use bot technology at home or work? Share your story with us on Facebook or Twitter.