Naturally Blogging

Speech recognition as a path to more efficient computing

Techie diversion today on something that I've been thinking about doing for a while. I'm writing this blog entry as my first experience with Dragon NaturallySpeaking, from Nuance Software. I have had a couple of colleagues sing its praises, and I've always been curious, so I finally ponied up the $99 at Staples and bought a copy of the Standard Edition. I've always been faster at editing than composing, and so I figured if I could just talk and then edit, I could get a whole lot more done more quickly!  The productivity implications could be stunning....

One of my key concerns was how well the software handled tech-speak and topics related to communications.  (LoL! I just had my dictation of "communications" translated as "comedic Haitians.")  This is a fairly simple program and it's been maturing for years. I saw earlier versions of this program 10 years ago and it worked fairly well, but it was somewhat costly and required a lot of training for it to really work. Now it's to a point where it's very inexpensive and with very little training (I spent a total of maybe 15 minutes) it's now doing a very good job of tracking my voice and translating the vast majority of it into the text that I want. This is the sort of technology that will make all sorts of computing much more accessible by the masses -- people who don't have years under their belts on on a keyboard doing battle with Windows or Unix.

It's really very exciting - I look at this product and I realize my wife, who is definitely technophobic and gets by via "accelerated hunt and peck" while fretting openly over why applications aren't organized in a way that makes sense, could benefit tremendously. Here's another funny translation -- "hunt and peck" first translated as "hunter pack" and then "hunchback" and then "hundred pack." I guess I need to do some more voice training.

Anyways, this sort of interface could make a huge difference in her life - not just because she's a medical professional (and there is a medical version of Dragon) but also in her role as family manager, trying to answer e-mails or writing letters or documents. For those without stressed carpal tunnels, writing on the computer will be far more easy when spoken than when typed using hunt and peck.

And for the savvy tech professional, natural language interfaces could hold great promise as productivity enhancers.  Many years ago (about 15, to my recollection) I first heard of a program that would do natural language recognition as a front end for network management. The concept was that you could give a voice command to change a configuration or open a dialogue or run a test, based on the data being presented to you. Makes a lot of sense, but the technology was way before its time. Plus, it was insanely expensive. In the face of the steady mantra of do more with less, this may be one of the technologies that allows professionals to accelerate their work tasks and processes.  This software does a great job of allowing you to call out menu options and select dialogue box options without touching the mouse or keyboard.

Watching how this software translates my voice, I can also foresee a whole new comedy genre based purely on how computers accidentally translate human voice. We can have great fun with the resulting translations and their altered meanings. Humor is often based on the unexpected results and surprises which come from a twisting of terms. Here's another one -- I can't get the translation software to recognize me saying the word "pun." It keeps wanting to translate as "open" or "upon." Clearly, this software itself doesn't have a sense of humor!

Well, it's time to get back to real work, but I'm having a lot of fun with this tool and it may just prove to be one of the more effective productivity improvers I've found in the last several years. In the end, I'd say that even though this was my first experience with the speech to text, it certainly saved at least half of my normal time for composition and editing. That means I can get a head start on my next post, where I will take another deeper look at comedic Haitians...

Anyone out there using this (or similar) software in front of their management tools today?  I'd like to hear about if you are....

Copyright © 2009 IDG Communications, Inc.

The 10 most powerful companies in enterprise networking 2022