http://blogs.msdn.com/sprague/archive/2004/10/22/246506.aspxIf you tried a speech recognition system a few years ago and weren't impressed, try again. The technology keeps getting better, slowly but surely, as this chart shows For many tasks, SR is already quite useful but the trend line shows how in not too manyen-USCommunityServer 2.1 SP1 (Build: 61025.2)http://blogs.msdn.com/sprague/archive/2004/10/22/246506.aspx#267528Sun, 21 Nov 2004 14:58:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:267528Chris BurdI don't know about that trend line. While I'm sure your calculations were totally legitimate, eyeballing the graph I estimate a one-third reduction in the error rate every three years. Assuming you're now at 5.5%, you don't get to 1% (your projection for 2011) until after 2017. If 0.5% turns out to be the practical minimum, you'll have to wait (well, not wait, really; research and innovate tirelessly) until after 2023. <br> <br>Chris Burd <br>cburd@katchword.fixtypo.ca <br>http://blogs.msdn.com/sprague/archive/2004/10/22/246506.aspx#9335412Sat, 17 Jan 2009 11:21:49 GMT91d46819-8472-40ad-a661-2c78acb4018c:9335412 Trend: screencasting - Frankwatching <p>PingBack from <a rel="nofollow" target="_new" href="http://www.frankwatching.com/archive/2005/03/21/trend-screencasting/">http://www.frankwatching.com/archive/2005/03/21/trend-screencasting/</a></p>