<br>I think it's a great idea, and I note that training simulations are a fairly<br>major market. More for corporate and professional training than for<br>kids or university students.<br><br>However, to make an avatar that was really useful as a surrogate
<br>teacher would require a really advanced AGI, I think. Except of<br>course if you're aiming at the preschool market -- a virtual<br>Barney or Big Bird shouldn't be hard to put together with<br>proto-toddler-level AGI technology!!!
<br><br>A near-term app of virtual agents is to serve as NPC's in training<br>simulations -- the so-called "serious games" market. For instance,<br>virtual criminals in police training sims, etc. ...<br><br>
-- Ben<br><br><br><br><div><span class="gmail_quote">
On 5/4/07, <b class="gmail_sendername">Jef Allbright</b> <<a href="mailto:jef@jefallbright.net" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">jef@jefallbright.net</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Ben -<br><br>Any comment on my suggestion of applying the technology to<br>motivational enhancement for learning, the virtual learning assistant<br>interacting with the student and providing intelligent feedback? Not<br>
so much a new idea as one that could be done better.<br><br>- Jef<br><br><br><br><br>On 5/4/07, Benjamin Goertzel <<a href="mailto:ben@goertzel.org" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">
ben@goertzel.org</a>> wrote:<br>><br>><br>><br>> Hi,<br>
><br>> > Well one application is obvious, considering the spread of the<br>> > pixel-sex trade in Second Life, and it wouldn't require a lot of<br>> > high-level intelligence to animate virtual prostitutes.
<br>><br>><br>> But I wonder if anyone would really like this, apart from the<br>> immediate novelty value? I don't regularly<br>> make use of prostitutes in the physical or virtual world,<br>> so I don't have a great understanding of the psychology of
<br>> people who do.... But isn't the fact that it's a HUMAN at<br>> the other end of the avatar important for the psychology<br>> of e-sex? At least, I think this would require a very convincing<br>> illusion of humanity. But of course, giving a convincing illusion
<br>> of humanity in that particular context might not be very hard....<br>><br>> A student of mine once wrote a chat bot that impersonated<br>> a hot and horny young 15 year old in online chat rooms. It<br>
> did very well and attracted a lot of email ;-p
<br>><br>> > But for more general applications, thinking along the lines of more<br>> > intelligently interactive PDAs should be a good bet. More<br>> > sophisticated phone answering, with intelligent message-taking
<br>> > (ensuring the important points are taken), prioritization and<br>> > forwarding; flexibly interactive appointment-taking on your behalf --<br>> > these are areas where we expect a human and are disappointed when we
<br>> > get a machine. If the machine agent can effectively represent the<br>> > specifics of its principle in such cases when the principle isn't<br>> > available, it should be a net positive.<br>
>
<br>> Yah, I see ... the famous virtual secretary, which according to the AI<br>> gurus of the 1960's was "right around the corner" ;-)<br>><br>> Presumably with a direct link into "Google Docs and Spreadsheets" +
<br>> Google Calendar or some such...<br>><br>> > Another area where personality counts, but doesn't require a high<br>> > level of intelligence, is in artificial pets.<br>><br>> That is quite possibly where we'll start ... I already have thought a
<br>> lot about that space though, which is why my question was about<br>> humanlike avatars specifically...<br>><br>> ><br>> > Similarly, but more suited to physical robotics, would be therapeutic<br>
> > devices that sense affect and respond with appropriate motivational<br>> > behavior.<br>> ><br>><br>> That is interesting. Ideally we would want to work with some<br>> physical device that automatically senses affect from the person's
<br>> body and voice, though.... Sensing affect from text is hard, which is one<br>> of the<br>> problems with email and chat communication. (Chat is better for<br>> affect than email, but achieves this at great cost in terms of loss
<br>> of subtle non-emotional content).<br>><br>> -- Ben<br>><br>> _______________________________________________<br>> extropy-chat mailing list<br>> <a href="mailto:extropy-chat@lists.extropy.org" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">
extropy-chat@lists.extropy.org
</a><br>> <a href="http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat</a><br>>
<br>><br>_______________________________________________<br>extropy-chat mailing list
<br><a href="mailto:extropy-chat@lists.extropy.org" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">extropy-chat@lists.extropy.org</a><br><a href="http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">
http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat</a>
<br></blockquote></div><br>