Eight Reasons Why Chatbot Technology Isn’t Ready for Commercial Use

In about three years, most people will be having more conversations with bots than with their spouses, according to Gartner Research. However, there are experts who believe that chatbots in their current state aren’t quite ready for serious business relationships. In this eWEEK slideshow, Stephen Hamrick, vice president of product management at SAP, explains how chatbots need to improve in order to be truly productive for employees and consumers. With these advancements, chatbots might earn the right to be involved in 50 percent to 60 percent of our conversations during the next few years. 

As artificial intelligence becomes more advanced, it will help chatbots to better understand people’s intent. With this ability, the bot could brainstorm new solutions that might not have been designed by the developer. Here’s an example:  If an employee was using a chatbot to book a conference room for a team of 10, a bot could find rooms with 10 chairs during the selected time. An AI-enhanced chatbot could learn from prior interactions that there are often last-minute guests to meetings and encourage the employee to book a room with 12 chairs to fit everyone comfortably.

Natural language processing still has a few challenges when it comes to turning complex requests into commands a chatbot can understand. Because chatbots can’t comprehend the nuances of intent or emotion from hearing human speech, they are not able to grasp the full meaning of many phrases or commands in a human capacity.

Chatbots now only know the correct responses to a narrow set of questions. They need more context and information to provide in-depth responses and solve more complex questions. This can be accomplished by securely integrating them with transactional systems and business applications, so they have the potential to provide a dynamic, actionable answer.

If it’s productivity, they are missing the mark. Yes, chatbots are engaging, but the parameters of when and how often to use them is becoming blurred. With instant messaging apps gaining traction and integrating with bots for everything from booking flights to ordering tacos, they currently serve more as a distraction to employees than a tool to increase productivity.

Humans and chatbots spend a lot of time talking to each other. But, due to a lack of standards and guidelines, there is currently no regulation for how chatbots interact with each other. In an enterprise, each department typically has its own chatbot, so at some point they will all need to migrate to one platform–which will be a painful process for some.

While chatbots in their current state might be limited, they are good at executing linear, repetitive tasks. These types of tasks include filing an expense report, uploading an image or registering an email address to a mailing list –all of them good candidates to be automated by chatbot because of their administrative qualities.

Part of the appeal of chatbots is to automate personalized messages to employees, but the reality is that this takes a lot of effort on the backend. Programming a chatbot to know everyone’s specific role and responsibilities would be easier if the chatbot was a tad smarter and could begin to learn and anticipate people’s jobs based on various factors such as experience, age and so on.

The AI within chatbots isn’t advanced enough to handle disgruntled employees or customer. They simply add fuel to the fire, because they are still learning how to interpret human emotions and don’t know when to pull in a human for help. But that doesn’t mean they need to be entirely cut out from the customer service equation. Some companies are testing out simple chatbots on the front lines of customer service to match user questions about a product or service with a defined set of knowledge-based articles.