I’ve been working on a project with Pat.ai/EC.ai for several months, and what those guys can do with linguistic understanding (and context) is pretty amazing.
Looking at a traditional customer experience in travel.
A traditional UI for example google’s multi-itinerary flight search form, has prescribed inputs
The customer enters data, select dates, and builds their itinerary. – comparing this to a dialogue with a travel agent the questions would be mostly the same. (“I’m interested in going to Singapore, I need to go via doha to attend a meeting but just for a day. Here’s my dates”)
When things start to change, this is when conversational processing offers a faster more natural interaction compared to a UI based traditional model. For example,
I need to change my dates to one week later.
(this could be 1 conversational request vs 3 UI date changes).
I need to visit Delhi after Doha for 2 nights.
(In the above UI that is lots of changes – give it a try).
Conversational UI Benefits
re-ordering, inserting and bulk changes are great opportunities for the UI to be more natural to the user.
There are plenty of interface opportunities to offer this time saving technology (chat bot, voice bot) on existing interactions and/or on new text only interfaces.
Natural language processing.
You could say any bot technology could achieve this, for the simple requests I would agree, but when things get more complex, it’s necessary to have context, and this is where NLU differentiates. Choosing a NLU engine with context can significantly improve the interaction and change a bot from dumb to really useful.
With the rise in home devices encouraging voice interaction, chat bots. I think conversational UI will see significant adoption, and will evolve from simple functions to complex ones, contextually aware and providing an experience comparable to a human today.