|
The conversational homunculus| old_uid | 13758 |
|---|
| title | The conversational homunculus |
|---|
| start_date | 2014/04/03 |
|---|
| schedule | 11h |
|---|
| online | no |
|---|
| location_info | Ampère |
|---|
| summary | A matter of fantasy, a matter of time, or a matter of fact?
The homunculus in a broad sense—a man-made creature with human characteristics and skills—has been a popular figure in literature for millennia. From the ancient myths of Philostephanus’ Pygmalion and Ovid’s Metamorphoses from the Antiquity, through the homunculus of Paracelsus from the Middle Ages, Goethe’s Faust and Mary Shelley’s Frankenstein from the Modern Era, and HAL 9000 of Kubrick’s and Clarke’s 2001 – A space Odyssey from contemporary times, it has been a strong theme in popular culture.
Today, the homunculus is an allegory for the long-term research goal of a number of researchers of human-machine communication, including myself. We strive, in the words of Justine Cassell, to create “a machine that acts human enough that we respond to it as we respond to another human”.
The research goal of creating a conversational homunculus – and artificial partner in conversation – may seem frivolous, but is well-founded from many perspectives, ranging from basic research in several fields to applied research, again in several fields. I will touch upon these motivations briefly, and argue that they are largely compatible, and can be grouped together to form one consolidated, strong motivation.
More importantly, I will address the question “is this at all a feasible goal”? Is it merely a pipe dream – an imaginary reward postulated merely to provide vain motivation? Is it just another problem that we are on track to solve? Or is it already solved – something we already have the knowledge and the resources to create? I will centre my talk around these questions, exemplifying whenever I can from my own research and from that of others, from KTH Speech, Music and Hearing and from elsewhere. |
|---|
| responsibles | Loevenbruck, Welby |
|---|
| |
|