During the last 4 months I have been working with a speech recognition software as an alternative way of typing text on the computer. My experiences has resulted in thoughts on the advantages and disadvantages of a these kind of systems. (I have used iDictate, but think the experiences would be the same regardless of system choice).
My experiences are both good and bad. The good thing about speech recognition software is that it allows me to make observations and write simultaneously. This is a really nice thing to be able to when doing research online, since the use of speech recognition software limits the shift between word processing and observations on a webpage. Over time, I believe it can become an advantage to use speech recognition software in regard to feeling immersed into the digital culture I am observing.
On the downside are the regularly word-training sessions. As a non-native English speaker my accent is neither British nor American. This challenges the flow of dictation to a very high degree. For about every fifth word I will have to start the word-training feature. My dictation becomes staccato and unnatural and the number of training sessions limits the work I can process in one day.
The most difficult challenge for me has been to adapt a new learning-style. I am a person who thinks as I am typing. Speech recognition software requires you to dictate in full sentences including grammatical punctuation. This really challenges my way of learning, thinking and developing ideas.
A colleague of mine once told me that it takes about 200 hours of training to be able to use a speech recognition software effectively. I believe the number is lower if you are a native English speaker. As I am locked-in to use speech recognition software in the future, this is a horrible amount of time, but it also fascinates me that we as humans still is so far away from making intelligent systems that can copy a human beings way of thinking and working.