Linked by Kroc Camen on Tue 28th Apr 2009 11:09 UTC
Podcasts More banal banter of the binary kind; Oracle buys Sun, Yahoo pulls the plug on Geocities, the Acer Revo nettop and much-offtopicness as we discuss the recently announced Windows 7 "Classic" XP mode.
Thread beginning with comment 360868
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE: podcasts...
by darknexus on Tue 28th Apr 2009 21:28 UTC in reply to "podcasts..."
darknexus
Member since:
2008-07-15

If we had speech recognition programs that good, we'd never need a keyboard or mouse again. Maybe in a few hundred years...

Reply Parent Score: 2

RE[2]: podcasts...
by Yamin on Tue 28th Apr 2009 21:40 in reply to "RE: podcasts..."
Yamin Member since:
2006-01-10

No way, I'll use a keyboard over voice recognition any day ;)

Seriously. I hate the automated phone system which ask you to say something.

Automated operation: "Say English to speak in English"
You have to say English.

Before it was "Press 1 for English"
You just press the number one.

It's so much less distracting to those around you. Imagine working in a word processor and having to say everything you type. The office environment would be screwed.

I don't know about you, but even if we had 100% perfect voice recognition, I'd still use the keyboard and mouse as they're just damn good, fast, precise, and QUIET.

Reply Parent Score: 2

RE[3]: podcasts...
by darknexus on Wed 29th Apr 2009 09:14 in reply to "RE[2]: podcasts..."
darknexus Member since:
2008-07-15

Those phone systems are a very good example of why our voice recognition technology is so bad at this point in time. They're trying to recognize all voices and, at the moment, that just isn't as practical as being trained to recognize one specific voice. To get even remotely reasonable results with voice recognition software at the moment, you need to train it to recognize your voice--that is what the reading procedure is doing when you first install one. It wouldn't be very effective for transcribing a podcast unless the podcast was being done by a single person. The only way you could get around this is to have voice recognition for each person in the podcast, and that would require the software to be installed on everyone's machines.
As for the phone systems themselves, I hate them, mostly because they absolutely seem to hate my voice. Don't know what it is about it, but they never seem to comprehend me no matter what phone I use. I'd also keep my keyboard over any voice recognition, just like I'd prefer if all phone systems still used number-based menus. Then again, if voice recognition ever reached 100%, I don't know for how much longer the keyboard would exist.
Anyway, back on topic... A very nice episode. Might want to fix the Garage Band crashes though. ;)

Reply Parent Score: 2