READ MORE

Everyone will have seen demos of hyper realistic digital humans holding deep conversations using AI and machine learning.  This technology is rapidly getting better and better, and more and more accessible.

Many users have high expectations, but the reality is that to deliver on this is still very ambitious and costly to develop and maintain.  Inbound virtual service agents based on chatbots would be the case-in-point.  This is a difficult use case because you don’t know who the person is, and the conversation could take many different paths.  These numerous dialog permutations need to be scripted, and kept up-to-date.  Unhappy paths have to be catered for, and the probability of a failed interaction is high.  That’s why many projects have stalled, or have been deployed to be less conversational by offering response buttons.

We are seeing the emergence of more narrow use cases for digital human technology.  Virtual service agents are becoming a more specialised offering, for example digital employees as part of a hybrid workforce, where the setting is more forgiving, and internally facing technology service desks.

Another way of controlling the use case is to zero in on outbound marketing, rather than the trickier inbound.  In this situation, the interaction is more controllable because you know who the person is you are interacting with, and you can control the direction of the discussion.  And people know the terms of engagement.  For example, in a survey they know that they are going to be asked questions and that they will give answers.