Utilizing AI for Chatbots

Artificial Intelligence and Machine Learning are becoming increasingly useful, and there are many applications that can benefit from them. As Julia Hu mentioned while presenting Lark, chatbots are changing user experience and user interaction. Julia demonstrated the capabilities of the artificial nurse, and mentioned that these chatbot nurses have replaced actual nurses, and are even reimbursable by insurance companies! Further advancements in AI and ML allow chatbots to become more interactive and natural to interact with, with natural language processing and context awareness.  

Redefining the user experience 

Chatbots have not always been easy for users to interact with. For example, automated customer service call centers that operate by listening for keywords can be cumbersome with words being identified incorrectly, or the inability to determine the purpose of the call at all. Even current digital assistance like Apple’s Siri or Amazon’s Alexa can have trouble recognizing context, and used to simply look for keywords. While chatbots and digital assistants continue to grow, they are not quite ready to completely replace humans, however they are on the path to mainstream acceptance. An article published by Forbes also gives four main advantages for chatbots over humans currently: “The chatbot never sleeps,” “The chatbot won’t make you wait,” “Personalizing the customer experience,” and “Chatbots make friends and build relationships.” [https://www.forbes.com/sites/shephyken/2017/07/15/ai-and-chatbots-are-transforming-the-customer-experience/#c0f363841f7b] The first two reasons given are self explanatory; for applications where chatbots are useful, they are always available since they do not rely on human interaction. The last two reasons are a bit more surprising to hear, since chatbots are not always elegant it is surprising to think they can improve customer experience and personalization. However, since chatbots are designed to collect information and process it accordingly, they can feed personal information to live support agents who can then use it to personalize their interaction with the user. In addition, chatbots are always available to reach out to users to develop brand loyalty: “Whether it’s a simple email or text on a customer’s birthday, or a quick check-in to ask if they are enjoying the brand’s product or service, chatbots are helping…This may sound counterintuitive, but the most sophisticated chatbots can provide a more human experience than an actual human.” [Forbes] Chatbots currently have great use cases, and as they become more advanced even more applications begin to open up to their use. 

Looking Ahead 

While customer support and user interaction are important aspects of chatbots to gain mainstream popularity, there are some applications currently in the works which could prove extremely useful. Negotiation is one such application, which is already making progress. According to an article from The Verge, Facebook’s FAIR lab has developed a new chatbot that can negotiate on behalf of the user: “The work is limited in scope, focusing only on a single negotiation scenario, but is a good proof of concept that points the way to building more powerful bots in the future. It also produced two particularly interesting outcomes. First, during a testing phase, the bots were often mistaken for humans; and second, that, without any human direction, the bots developed nuanced negotiation strategies.” [https://www.theverge.com/2017/6/14/15799068/chatbot-negotiations-ai-facebook-fair] Using a dataset of recorded negotiations, the neural network behind the chatbot was trained to negotiate and incorporate what they call “dialogue rollouts:” ” Essentially, this means asking the bots to think ahead; to simulate how future negotiations might go and pick the best course of action to get what they want.” [The Verge] While the article goes on to specify that this bot and research are more of a step in the right direction than a breakthrough, there are chatbots that exist today that can negotiate for consumers! The article specifies two examples, one is a bot that will negotiate with Comcast by exploiting the fact that Comcast customer service reps follow a strict script. The second example is a bot called DoNotPay that has fought over 160,000 parking tickets in New York and London. [The Verge]  

Conclusion 

Chatbots have developed significantly over the years, from irritating phone trees that most users wanted nothing to do with, to context-aware digital assistants which are now being purchased regularly. As AI and machine learning become more powerful, they can be applied towards the many aspects that would make up chatbots. It is feasible to think that within a few years, everyone could have their own digital assistant that could negotiate for them, handle their schedule, and even handle finances!

Users who have LIKED this post:

•