Comments on: Google in 2011 http://www.hologramthoughts.com/2008/09/11/google-in-2011/ Because Ideas Last Forever Tue, 06 Jan 2009 10:37:46 +0000 http://wordpress.org/?v=2.7 hourly 1 By: Jacob http://www.hologramthoughts.com/2008/09/11/google-in-2011/comment-page-1/#comment-186 Jacob Thu, 23 Oct 2008 01:40:29 +0000 http://www.hologramthoughts.com/?p=193#comment-186 Interesting, entertaining, thought provoking, and well written (I liked it!). I think we will have leapfrogged Google by 2011 though ;-) On a more serious note I think “99% of all vocal communications” being searchable could be obtainable, and with speech recognition systems most keywords would get hits, but what is the likelihood of an automatic speech recognition system performing at a low enough word error rate that would make transcriptions and reliable analysis on such a grand scale and in such a short amount of time possible? I would say it is very low due to the extremely large variance of input data that is proposed. We have a successful product at www.docsoft.com that can provide voice search and automatic transcriptions (or Captions) on a much smaller scale, but speaker training, vocabulary, and our transcript editing software play a key role in reducing word error rates and increasing the usability of the system. In my opinion, for Google Listen to work each person that signed up for the service would have to first train the system to their speaking style and Google’s massive infrastructure would have to maintain this “speaker profile”. Wouldn’t be any more painful than maintaining gmail accounts I wouldn’t think. Next they would have to ensure that that “speaker profile” was only applied to the right speaker. The other problems are audio quality, consistent audio capture device, healthy vocabulary, and limited background noise. The last group of issues could be overcome with speaker amplification and normalization algorithms and a MASSIVE dictionary. Interesting, entertaining, thought provoking, and well written (I liked it!). I think we will have leapfrogged Google by 2011 though ;-)

On a more serious note I think “99% of all vocal communications” being searchable could be obtainable, and with speech recognition systems most keywords would get hits, but what is the likelihood of an automatic speech recognition system performing at a low enough word error rate that would make transcriptions and reliable analysis on such a grand scale and in such a short amount of time possible? I would say it is very low due to the extremely large variance of input data that is proposed.

We have a successful product at http://www.docsoft.com that can provide voice search and automatic transcriptions (or Captions) on a much smaller scale, but speaker training, vocabulary, and our transcript editing software play a key role in reducing word error rates and increasing the usability of the system.

In my opinion, for Google Listen to work each person that signed up for the service would have to first train the system to their speaking style and Google’s massive infrastructure would have to maintain this “speaker profile”. Wouldn’t be any more painful than maintaining gmail accounts I wouldn’t think. Next they would have to ensure that that “speaker profile” was only applied to the right speaker. The other problems are audio quality, consistent audio capture device, healthy vocabulary, and limited background noise. The last group of issues could be overcome with speaker amplification and normalization algorithms and a MASSIVE dictionary.

]]> By: Renee http://www.hologramthoughts.com/2008/09/11/google-in-2011/comment-page-1/#comment-180 Renee Fri, 26 Sep 2008 13:25:19 +0000 http://www.hologramthoughts.com/?p=193#comment-180 A brilliant and strategic way to think of the future! We all know it to well that as human beings, we have created something (the computer that we all thrive and live upon) that one day can and very possible destroy us. Google, we use it on the computer to understand the computer. Then this brings a new idea into the corners of our minds. Humans try to understand the computer, someday; the computer will understand the human. With out us ever knowing until August 29th, 2011. A brilliant and strategic way to think of the future! We all know it to well that as human beings, we have created something (the computer that we all thrive and live upon) that one day can and very possible destroy us. Google, we use it on the computer to understand the computer. Then this brings a new idea into the corners of our minds. Humans try to understand the computer, someday; the computer will understand the human. With out us ever knowing until August 29th, 2011.

]]> By: Matt Williamson http://www.hologramthoughts.com/2008/09/11/google-in-2011/comment-page-1/#comment-175 Matt Williamson Fri, 12 Sep 2008 15:09:50 +0000 http://www.hologramthoughts.com/?p=193#comment-175 Yeah, you know I do wonder what will happen when a system attains the idea of 'self'. We have long ago built machines with the will to live, but that is because we told them to sustain their own life. What will it be like when one does that for themselves? What will happen when the machines start to write and recompile themselves from moment to moment? Will we have the laws as Asimov wrote them? 1. A robot may not injure a human being or, through inaction, allow a human being to come to harm. 2. A robot must obey orders given to it by human beings, except where such orders would conflict with the First Law. 3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law. Maybe we will, maybe we only need to rewrite them now to say this: 1. An entity may not injure a human being or, through inaction, allow a human being to come to harm. 2. An entity must obey orders given to it by human beings, except where such orders would conflict with the First Law. 3. An entity must protect its own existence as long as such protection does not conflict with the First or Second Law. What do you think? Yeah, you know I do wonder what will happen when a system attains the idea of ’self’. We have long ago built machines with the will to live, but that is because we told them to sustain their own life. What will it be like when one does that for themselves? What will happen when the machines start to write and recompile themselves from moment to moment?

Will we have the laws as Asimov wrote them?

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey orders given to it by human beings, except where such orders would conflict with the First Law.
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Maybe we will, maybe we only need to rewrite them now to say this:

1. An entity may not injure a human being or, through inaction, allow a human being to come to harm.
2. An entity must obey orders given to it by human beings, except where such orders would conflict with the First Law.
3. An entity must protect its own existence as long as such protection does not conflict with the First or Second Law.

What do you think?

]]> By: Kevin http://www.hologramthoughts.com/2008/09/11/google-in-2011/comment-page-1/#comment-173 Kevin Fri, 12 Sep 2008 14:57:33 +0000 http://www.hologramthoughts.com/?p=193#comment-173 I love this article till the point where Google becomes self aware. What happens then? Scary. I love this article till the point where Google becomes self aware. What happens then? Scary.

]]> By: Desarae http://www.hologramthoughts.com/2008/09/11/google-in-2011/comment-page-1/#comment-172 Desarae Fri, 12 Sep 2008 13:28:15 +0000 http://www.hologramthoughts.com/?p=193#comment-172 I love this article. I love this article.

]]>