dangerous_dom Wrote: 
> I really hate voice command stuff. I just feel like a complete idiot and
> nerd using it. I really dont want to talk to electronic stuff, why are
> some people to obsessed with it? I guess because it's all futuristic
> like star trek. No, not for me.
I have a friend who is a musician with severe repetitive stress
injuries from playing violin and piano. She was working on her master's
degree when her hands kinda gave out. I offered to help her by typing
some of her thesis (it was mixed English/German and not so easy to just
hand to a secretary) but she opted instead to get a grant and get a
computer with voice response. OK, this was probably 8 years ago, but it
sucked the big one. Every time she said something slightly wrong, it
would type it wrong. Then she's say "oh, shit" so then it would type
that. You get the point.

There's no good syntactic way to give out of band or meta information.
We as good listeners running a syntactic and semantic engine in
parallel understand when someone goes back and corrects something, but
computers don't have nearly the horsepower we have and software isn't
there yet anyways. Not by a long shot.

When I was in undergraduate school lo those 30 mumble years ago we were
told voice recognition was around the corner, 10 or 15 years or so. I
stayed in the sidelines of academia for another 15 years and each year
got the CSC reading lists and bought all the books. Every few years
they'd push that estimate out. If anything, all that's happened in the
mean time is we have realized what a large goal that really was and
gained new respect for how amazing the human brain is.

Now, I'm far even from the outskirts of academia these days, but I do
subscribe to a kind of clipping service for SciAm. If there had been a
breakthrough in voice recognition in the last few years, I think it
would have knocked on my mailbox, either through the SciAm feed or some
other way. Nobody's come knocking, so I doubt there's any breakthrough.

Without a breakthrough, it's like Dangerous Dom says. Won't work in a
room where the music is already playing. Won't work for different
people without retraining. Won't work if you have a cold. etc.

Someone did a nice job with photoshop, and someone's looking for
investors (I think that's the main thing). I think it probably works
marginally enough to be a gadget. They have 3 voice recognition
products. Polaroid is one of the backers. I'm sure it's not a total
crock. It's just that the distance between "kinda works" and "works
well enough that it's like talking to my friend" is so vast, most of
the public won't realize it either until they've bought into it.

To give you an idea, my lawyer's answering system asks you to speak the
name of the person you want to speak to. I say his name, it connects me
to him.

My cell phone can dial for me if I say the name of the person I want to
speak to.

Each of these examples works with a vocabulary of a few dozen words.
How many of you have a few dozen songs in your library?

Think about saying "I want to listen to a version of All along the
Watchtower, not the hendrix version but the original dylan version, but
not from a live concert".

It's a long way from saying "Call Home" to your cell phone.


-- 
Michaelwagner
------------------------------------------------------------------------
Michaelwagner's Profile: http://forums.slimdevices.com/member.php?userid=428
View this thread: http://forums.slimdevices.com/showthread.php?t=25541

_______________________________________________
discuss mailing list
[email protected]
http://lists.slimdevices.com/lists/listinfo/discuss

Reply via email to