The Top 5 Misconceptions You Need to Know About Chatbots

By Kristi Colleran - October 10, 2017


 

When chatbots re-emerged in the Spring of 2016, information about them was limited. But as the technology has advanced and become more mainstream, so has the availability of information…and some fallacies, too.

Let’s take five and clear up five of the most common misconceptions!

Misconception #1: All Chatbots Use Artificial Intelligence

Fact: A chatbot may…or may not…use artificial intelligence.

There are 3 basic types of bots:

1) Scripted chatbots rely on a rules-based conversation that follows predetermined paths. At each step in the conversation the user must pick from explicit options typically presented via a menu and/or buttons.

2) Intelligent chatbots are built with artificial intelligence, in particular the Natural Language Processing (NLP) technologies of Natural Language Understanding (NLU) and Natural Language Generation (NLG). NLU interprets what the user says and maps it to a computer representation. NLG produces a natural language response that can be understood by the user. Currently the content of the responses aren’t created by the technology but typically written and/or developed by humans.

There are two sub-types of intelligent chatbots:

  • God Bots which include Alexa, Siri, Google Assistant and Cortana. These bots possess general knowledge across a wide range of topics.
  • Purpose-Driven Bots in contrast to God Bots have a narrow focus. Purpose-driven bots are designed and developed to provide information and/or perform tasks within a tightly defined scope.

In the not so distant future, you may be asking your favorite god bot to coordinate with purpose-driven bots to accomplish tasks on your behalf such as scheduling meetings, booking travel, ordering pizza and sending flowers!

3) Hybrid chatbots use scripted conversation supplemented by NLP to support free-form texting. The menus and/or buttons of the scripted conversation help guide users and the artificial intelligence ensures the chatbot can handle a broader range of texting interactions such as when the user deviates from the script by free-form texting.

Misconception #2: Chatbots Can Run on any Messaging App

Fact: Chatbots can run on many different messaging apps but not all.

WhatsApp, the dominant world leader in messaging apps, currently prohibits chatbots in their Terms of Service. But, stay tuned this may change. WhatsApp recently announced they are verifying business accounts and exploring ways for businesses to message with their customers.

Chatbots may not be allowed on WhatsApp yet, but they can run on many messaging apps including but not limited to: Messenger, SMS, iMessage, Twitter, Telegram, Viber, WeChat, Kik, Slack, Cisco Spark and Microsoft Teams. In addition to messaging apps, Chatbots can also run on a website or within another type of app such as for banking or shopping.

While chatbots can run on many different messaging apps and even on websites and in apps, the technical requirements for each of these differ. And the user experience will also differ depending upon the functionality supported.

Misconception #3: Chatbots Can Learn Entirely on Their Own

Fact: For most organizations, intelligent and hybrid chatbots need to be developed and trained by humans.

True machine learning currently requires massive amounts of data that very few organizations have and even fewer have the capability to use for machine learning. The human training of chatbots is known as supervised learning. Once a chatbot is live and responding to users on it’s own based on what it has been developed and trained to do, it can learn on it’s own to a certain degree within a set of parameters. However, you may not always like what it learns! So a bot’s unsupervised learning must be reviewed by humans and corrected as needed.

Misconception #4: Chatbots Can Replace Human Staff

Fact: While chatbots can handle basic questions and tasks, humans are still needed to address complex questions and issues.

“…If you look at almost every other tool that has ever been created, our tools tend to be most valuable when they’re amplifying us, when they’re extending our reach, when they’re increasing our strength, when they’re allowing us to do things that we can’t do by ourselves as human beings. That’s really the way that we need to be thinking about AI as well, and to the extent that we actually call it augmented intelligence, not artificial intelligence.”

Rob High, VP and CTO — IBM Watson

Here are three ways chatbots can augment human teams:

1) Always Available — Would you like to expand your team’s availability but face budget and staffing contraints? Good news! Chatbots are economical and can work 24 hours a day, 7 days a week, 365 days a year. And they are always on time, pleasant and never take a personal day.

2) Ability to Scale — Do you find it challenging to scale your team during peak times? Or you have an opportunity to expand or grow your business but need to support it? Chatbots facilitate scaling by augmenting your human team. You can use chatbots as the first point of contact to handle basic questions. This frees up your human agents to address more complex questions and issues. If the chatbot is unable to address the customer’s need, it can invite a human member of the team to the conversation.

Adding a chatbot to your team means more customers served! And with the basic questions handled, your team is freed up to address more complex questions and issues.

3) Consistent, High Quality Service — Your team likely answers the same questions over and over again. Chatbots are well suited for answering your customers’ frequently asked questions and they never tire of them, so each and every one of your customers gets the same high quality service.

Misconception #5: Users Shouldn’t Be Told They’re Talking to a Bot

Fact: No transparency, no trust.

While chatbots can mimic human conversation and interact with humans, they are not human. In a study of thousands of individuals, CEB (now Gartner) found that transparency ranked as the highest priority for consumers and it also has the largest impact on loyalty. Trying to pass off a chatbot as a human — no matter how well-intentioned — can damage the customer relationships you’ve worked so hard to build.

Almost every customer experience expert agrees: You should always be honest with your customers about whether they are interacting with a bot or not. As Beyond Philosophy’s Michael Lowenstein puts it, “If organizations are not transparent about how and where chatbots are being used with customers, and that the chatbots are offering responses only as programmed, trust could be very seriously impaired, even destroyed.”

 


About the Author

Kristi Colleran is the Co-Founder of Sciensio and heads up AI Interaction Design for EventBots by Sciensio. You can also follow her on Twitter.

3phoneDriveEventImage.png

Case Study

Drive Event

Deploying Sciensio's Eventbots solution to improve and enhance a major industry trade show.

Get Case Study
Comments

We promise that we won't SPAM you.