As far as voice or speech recognition is concerned, it will never be able to replace a human being. It makes too many mistakes and it will cost double to then have a medical transcriptions go back to review and edit the reports. Since medical transcription has to be 100% accurate, there is no room for the slightest of errors. Also, many doctors don’t have the time to dictate, proofread, and edit the work to make sure it is 100% accurate. It is too time consuming enough to see patients and just dictate them to be dictated into reports. Speech recognition works well with e-mails and basic word documents, but it is too complex for medical transcription. Speech recognition may work with really small companies as doctors do their own dictating in-house and see very little patients, which gives them the time to carefully edit and use the voice recognition system. In short, your future as a medical transcriptionist is safe.
Examples of voice recognition errors: