How do they work?

The rudimentary processes involved in how chatbots work are generally categorized in two tiers: either “stateful” or “stateless”.

“Stateful” chatbots run on the idea that their interactions with humans are based on previous interactions with other people, while “stateless” chatbots are designed to engage in “conversations” without references to past interactions – IE, to treat every interaction as though it were a new one.

At their most basic, chatbots are structured with a highly scripted base that allows them to give answers to frequently asked questions or perform repetitive and rudimentary tasks like asking for a subscription code or account number to verify accounts or identities.

Developments in the field have seen to the rise of various chatbot types, and these include scripted or quick reply chatbots (chatbots that operate on a set of predefined questions that follow a hierarchical decision workflow), keyword recognition-based chatbots (chatbots that’re designed to “listen” for particular keywords and provide an appropriate response to questions), hybrid chatbots (chatbots that integrate aspects of scripted and keyword-based bots) and voice-enabled chatbots (chatbots that utilize text-to-speech and voice recognition tech in doing what they are designed to do).

The contextual chatbot type – one that works with an artificial intelligence matrix and machine learning systems – is the type that has roused the attention of many, in the way these are designed to have a “memory” of interactions with people, then use what they “remember” when they are “talking” to others in relation to keywords, predefined questions and other sources.

Basically, these chatbots “learn” and develop a “personality” as they interact with people more and more.