The chatbot movement has not been straightforward. I think we can all agree that some of the blitz marketing that appeared around the globe telling us how smooth human chatbot communication was going to be, did not help the industry set realistic expectations. Nor did that marking foster the kind of incremental improvements that other technology releases benefit from.
Like the advent of GPRS, Edge and mobile 3G technology, the marketing bods got hold of a vision and sold it to us, before the bytes started flying. As a result the world was universally disappointed with mobile data until the advent of 4G.
So with chatbots, the marketing teams knew what HAL from 2001 a space odyssey sounded like and used it to set the wrong expectations entirely with AI chatbot customers globally. In reality, chatbot maturity in 2019 is perhaps closer to Mobile Data 3G capabilities. It has it's uses and can really provide a few powerful services, but we still have a long journey to travel to meet customer expectations.
I expect to see some real chatbot traction when the problem of training material is resolved and when we can start to build relevant ontologies. Whatever the milestone, those of us who have been building chatbots from 2014 onwards can justifiably feel aggrieved at having been given a hole and a shovel and asked to build a mountain.