A patent filed by Google engineers describes how chatbots will one day show empathy for the user -- if they are not already doing so. The patent uses the word “chatbot” to refer to interactive or virtual assistants on desktop or mobile devices.
Despite efforts to make chatbots seem more human-like, they may still seem unnatural or awkward because they do not keep track of users' emotional states.
A chatbot may greet a user with the statement "hello, how are you today?" If the user tells the chatbot "I feel lousy today," it would detect the negative state expressed and may select an appropriate response, such as "I'm sorry to hear that."
This is where it would end unless artificial intelligence and natural-language processing are used to help it learn -- not just on devices, but also in search engine queries and websites.
The idea is for chatbots to “achieve greater social grace by tracking users' states and providing corresponding dialog,” according to the patent description granted by the United States Patent and Trademark Office on Dec. 24, 2019.
The chatbot will determine the “state” of the person by the words or sequence of words being used. The patent application describes how the state of the user can be identified in the person’s natural-language output. This could also identify types of spam bots.
“An indication of the state expressed by the user may be stored in memory for future use by the chatbot,” according to the patent. “It may then be determined, e.g., by the chatbot based on various signals, that a second session between the user and the chatbot is underway.”
The developers of this patent believe that the more human a chatbot appears, the more likely that a user will converse with it in the future.
The more the chatbot is used, the more it learns about the user and their lifestyle and interactions. It becomes more intelligent and can recommend resources and become increasingly useful.