About VUI vs CUI (again)…
you provocatively :) ask:
There are a dozen ways to order a coffee. Why do dumb bots only allow one?
The short answer to your question is simple:
it’s because these “dumb bots” are probably bad designed, probably hard-coded by software programmers (traditional computer language developers that realize bots with NodeJS/Python/etc.).
SW Developers (and I’m a developer me too) don’t have the correct mindset to realize smart conversational experiences and probably they still practice wrong tools to build-up innovative, or simply smart conversational flows (see my old notes, e.g. , , , etc.). I explain better in a moment, but beforehand let me point out the your article contains great graphics and insights. Thanks.
Hi already commented Mikail’s (great) article, me too, and … I’m in the middle.
On one hand, I’m pretty sure that in near future chatbots will be done through a fully conversational UX (maybe without a Graphical User Interface but instead #voicefirst, with a prevalent voice-devices based interface, and, as chatbot developers,
we need to study & deepen the humans-to-humans usual language interaction behavioral patterns of common natural language understanding
BTW, that’s the reason of my comment above mentioned and my frustration with widgets, webviews and, in general, with what I call the click&point paradigma.
On the other hand, I do not like too much the RAN (Random Access Navigation) definition, because I associate it in my mind with stateless intent/entities extractors (with machine learning) , now excessively popular :-)
Viceversa, RAN is a real requirement if you mean a way to back the user to a (business application) contextual (work)flow, and, paradoxically, but please note, there is no so much AI on that,
RAN is simply a smart processing & memory of user conversational acts (within the business application)