AI in 2019: Are chatbots the future?
Mention the phrase ‘automated phone system’ and no rationale person thinks of the free exchange of information and ideas. The limits are due to framework constraints, a lack of critical thinking and a proxy set that can only answer the most basic of questions. Anything more complicated than this requires human intervention because AI is in its embryonic state. However, if you need information, and because your options are limited, then you must enter into the push-button phone matrix that not even Keanu Reeves can crack! But, what if there was a system in place that knew the information completely. A system with intuitive, instant recall that could respond to unique and person-specific inquiries – in any language – with ease. No matter how poorly or accurately a question was asked. Welcome to Chatbot Nirvana – which doesn’t exist….at least not yet!!!
The problem with automated phone systems, and chatbots, in general, are the constraints of system understanding and the complete lack of functional flexibility when it comes to responses. The standard ‘pressing of buttons’ to move through a system is one that we are used to as long as the information is basic and is backed up by the ability to hit zero and get a person. However, the need to automate and streamline how and when people can access information is a long-standing, key problem that requires vision, focus, and flexibility. It is an issue that begs the question: Are Chatbots – even in Federal institutions – the future of automated phone systems?
At Viderity we feel the answer is a resounding yes and we are looking to a future with a more useful and complete Chatbot. One that can truly ‘know’ a system and, more critically, respond to requests along multiple lines of reasoning and informational bundling. Essentially, an automated phone system that thinks on its feet, responds to inquiry and has a more complete understanding of the issues which can arise with questions. But can a Chatbot be made to think in such a way as to not only mimic a free-flowing exchange of ideas but, more importantly, to give intelligent and reasonable answers?
At Viderity we feel that the Endgame Chatbots will do just that. In order to get to that level of specificity, we need to start with a more fundamental challenge. A system that, at first, simply listens to a question and not only hears what is being asked but can also link specific issues to this inquiry. Possibly even re-phrasing a question so that it more accurately fits the design paradigm of the program. The foundation of these systems is found in their linear construct which allows for simple responses. The next phase of which is a more tailored response that moves us closer to a functioning AI.
The foundational concept of a complete Chatbot is here and what was once considered impossible is rapidly becoming possible. Informationally dense institutions, such as the Federal Government and other huge databases, are seeking more cost-effective and efficient ways to transfer information. The initial goal, then, is to become more adept at identifying types of questions and responding in a more fluid manner. In addition to this construct, the ultimate challenge is to have a system that gains enough insight from inquiry to supply both relevant and sufficient information. A system that, in fact, cuts through the bureaucracy and provides appropriate and helpful responses by using true AI.
A system like this will require years of intense and diligent work in order to even begin to come close to the type of ‘AI thinking’ we are seeking. The groundwork for this type of system is already in motion at Viderity. We are future-focused on building a solid, relevant system that thinks its way through a conversation and tailors its responses accordingly. We feel that AI and the search for a functioning form of Chatbot is within the realm of possibility and we are working diligently towards that reality.