The Future of Transcription
Way With Words
The Future of Transcription
What is the future of transcription? Tell us your thoughts.
Did you know that the first developments in speech recognition were carried out more than 50 years before the modern computer blasted on to the scene? The man who brought us the telephone, Alexander Graham Bell, started experimenting with transmitting speech and was driven to do this as his wife was deaf. He wanted to create a device that would take words and turn them into a picture that a deaf person could then understand. He did actually land up producing images of sounds, but his wife couldn’t understand them. From this line of research he landed up inventing the telephone.
So we move on a few years. We all know that technology is moving faster than the speed of light, and what’s relevant today is not necessarily relevant tomorrow.
How will technology change the future of transcription?
The idea of talking to a computer and actually getting an answer, or a transcript of what you’re saying, was well the stuff of movies! How things have changed. You talk in to your smart phone and it actually talks back, finds something for you or even sends a message to someone. Granted it doesn’t get it right a lot of the time and it is rather amusing, but the future is here.
There has been a lot of press coverage recently about the improvements coming our way in terms of voice to text programmes and applications. The major focus has been on voice and the understanding of natural language. This is where technology has failed up until now. Currently there are programmes out there where you have to ‘train’ it to understand your natural way of talking.
Sophisticated voice technology is already commonplace in call centres. Users navigate through menus and they are able to pick up irate customers which are then passed on to a customer service rep.
Accuracy is a key thing moving forward for voice to text, but how accurate can it get?
Consider things like slang, dialects, accents and background noise. Will a computer programme ever fully be able to fully understand these factors?
In our opinion, although technology is improving rapidly and that transcription in its traditional sense will change, there will never be 100% accuracy. Consider things like legal hearings, doctor’s notes, research, news amongst others. Getting it wrong is not an option as 100% accuracy is needed, not 90%. There will always be the need for transcribers. Why? This will ensure the level of accuracy that is required for highly sensitive recordings. We believe that technology and humans will need to be combined. A transcribers job might change over time, but there will very definitely still be a need for the human element.
What are your thoughts on how transcription will change in the future?