Slashdot Mirror
The Virtual Teacher
Roland Piquepaille writes "Researchers from Illinois and Florida are developing a networking system which will create virtual representations of real people to improve our knowledge. They will use artificial intelligence and natural language processing software to enable us to interact with these avatars. The goal of the project, sponsored by the National Science Foundation (NSF), is to give us the possibility to interact with these virtual representations as if they were the actual person, complete with the ability to understand and answer questions. We should see the results at the beginning of 2008 — if the researchers succeed."
Summary of article (IMHO):
Guess what. In the future, we will create avatars that will look and sound just like real people. To see how far along we are, take a look at some of these avatars
(decent graphical images, which not an AI make)
There's some money from the NSF to continue this effort.
Really smart guys are working on ways to get realistic sounding voices and
mannerisms.
We're using motion capture too.
Is this commercially viable? We'll have to see.
First of all, this virtual avatar stuff is icing on the cake. The first (easiest) Turing test implementation does not require body language, realistic avatars, or
mannerisms, and we are still YEARS from that. What makes anyone think that this is remotely possible by next year?
P.S. The submitter's name seems familiar. Where have I heard it before?
Rhymes that keep their secrets will unfold behind the clouds.There upon the rainbow is the answer to a neverending story
Allergies are blocking my nose, but I can still smell the stench of sensationalism with this one.
'Every story, if continued long enough, ends in death.' --Ernest Hemingway
I doubt we will see any meaningful result soon on this field. First because AI is not at this advanced point yet (in order to provide improvisation and interactivity), and the use of pre-recorded images and sounds cannot cope with the ingenuity and the unpredictability of the students. Second because human beings, although very sympathetic to the idea of talking heads (just look at clippy), in the long run get bored of it.
On the other side, I already had a couple of teachers that were completely inept to their jobs, following scripts just like a computer program with pre-recorded footage would. But computers are very unlikely to come even near to advantages a real good teacher can provide.
P.S. The submitter's name seems familiar. Where have I heard it before?
Roland submits dozens upon dozens of stories to Slashdot. "Vacuous" is pretty accurate; I've repeatedly heard Slashdotters ask for the ability to filter his crap like we used to be able to filter Jon Katz.
The primary difference is that while Katz was batshit insane and overimpressed with himself, Roland simply states the obvious and inane in his "article", and then submits it to slashdot. He regularly does little more than quote sections of the article and supply obvious commentary...and watches the hit counts roll in.
There are two remarkable facts: one, that there isn't better content in the submission queue, and two: I don't think I've ever seen comments posted supporting him. Hilariously, on the rare occasion he does post, he's moderated down so fast, he must be on the level of the GNAA people in terms of Karma.
Since hits support Roland, I'd suggest slashdotters tag his stuff "boycottroland"
Please help metamoderate.
I know that a lot of people here at Slashdot have a burning hatred for Roland Piquepaille. But I find it confusing. The articles he submits are always very interesting, especially for those of us interested in science. They're the epitome of "news for nerds".
The problem is that he submits interesting articles which are linked via his blog. So we read a blog story about a blog story about something interesting.
He forces slashdotters to go through his site (racking up hits and advertising dollars) to read the actual interesting content...
Please help metamoderate.
I've just added *.primidi.com (Roland's blog) to my adblock filters. I obviously never go there intentionally, but I've clicked on his links a few times without checking the submitter name. I suggest you all do the same.
Unfortunately, the ZDnet article linked is also written by our buddy Roland. I do wish that Slashdot would give me a way to avoid his crap (or, y'know, just not approve his stories). Either way, I think the boycottroland tag will help me in the future.