Over at Forrester, Diana Clarkson’s blog brought up an interesting topic – What should avatars look like for virtual agents?

http://blogs.forrester.com/diane_clarkson/11-03-09-what_should_a_virtual_agent_look_like

As you can see, or not see, we don’t incorporate visual avatars with our contentAI virtual agents, characters or mLearning applications (See:  ESLai.com)

Why not?  Especially since the Principals of the company come from a deep motion picture and CGI background?

Basically, the technology isn’t good enough right now for real-time, on the fly,  virtual avatars.  Yes, we could pre-render much of them and pull off some fancy tricks to cover text-to-speech realtime adjustments — But, we don’t find this to be satisfying for the end user.  Frequently, the end user doesn’t want to turn up the volume and listen – Nor do they want to watch…when reading allows them to quickly scan/read or look away and return to the content on their own schedule.

As one comment on the Forrester blog mentions, using the dialogue itself to paint the picture for the end user is consistent with our approach.  The end user’s imagination fills in many blanks.  As we review the (anonymous) chat logs and see 10-20 minute sessions, we know the end user is “buying into” the experience without the aid of a cartoonish animation.

Our preference is for incorporating a video, audio or still image within the first step of a chat session to “establish” the experience, then, let the User create their own “vision” from there.  This also personalizes the user experience more than if we provided a simplistic visual.

We did like the follow up COMMENT from Clarkson on the post:

“I think we’ll see a lot of interesting developments in the virtual agent space with mobile in the coming year.”

In our opinion, that’s starting today!  Right here.