Many folks have seen the YouTube video above, an Apple demonstration from 1987 titled Knowledge Navigator. Now that we’re in 2018, have we achieved this level of interaction?
DisEmbodied Location Conversational Agents (DELCA) are intelligent agents that interact with their environment (including us). This could be a web site with a computer avatar or a conversant robot.
There have been some interesting embodied agents over the years. This includes MAKEbot from Make back in 2006. The U.S. military has been using an agent on their website – Sgt. Star – for awhile. The FBI and CIA use a similar system to the Army. Some shopping websites have used avatars also.
Agents are not above running into controversy – Sgt. Star was cited for failure to release records via a FOIA request back in 2014 from EFF.
As the day comes closer that Google and Apple look to make Alexa and Siri visual avatars, able to have real-time conversational discussions, are we heading to a future envisioned by Apple last century?
Hopefully recent research will give us an idea. There is a good group of folks writing cogent papers on the subject:
- MIT – Task-Oriented Collaboration with Embodied Agents in Virtual Worlds
- Book – Emotion and personality in a conversational agent
- Journal – Building character for artificial conversational agents: Ethos, ethics, believability, and credibility
- . . . and other articles
There are likely a few we’re missing in this brief overview, post’em up in the comments!