Actually, voice recognition does not necessarily have to put transcriptionists out of work. That is one benefit many people will see immediately, but the best voice recognition software on the market today is still only 99% accurate with a good amount of voice training involved. And people using the software don’t always have the time to constantly train their software. Now 99% might seem pretty good and it is, but that 1% can be a big deal. Say for example a doctor is dictating a medication dosage. He says 15 mg of Medicine X but the speech recognition software hears 50 mg. And maybe Medicine X is lethal beyond 30 mg. Now you have a big problem because 15 and 50 are somewhat similar in sound. This is where transcriptionists come back into the picture and save the day.
Transcriptionists can easily take on the role of QA (Quality Assurance). They can listen through the dictation created by speech recognition and follow along with the typed document, and make sure everything is correct (so you don’t give someone 50 gm instead of 15 mg for example). Using 3rd Party Correction features they can also make corrections that will reflect on the original user’s voice profile. So the document is being proofread and the voice profile is being trained all at the same time with the help of a transcriptionist working as QA.
So, many companies will at first see voice recognition as a big money saver because they think they will be able to eliminate transcriptionists. However, they will quickly realize that transcriptionists are still an essential part of creating finished documents even with voice recognition available.