Will AI Disrupt the Transcription Industry and Replace Human Transcriptionists?

Voice recognition accuracy is improving each day and voice is an important area for most of the top technology companies in the world including Google, Amazon, Apple, Facebook, Baidu and Youtube. New applications are being introduced in the market on a daily basis and it’s baffling to witness the speed of innovation.

Not long ago Apple introduced the ‘Clips’ app that does real time transcription with an astounding accuracy and this week, YouTube announced that it’s bringing automatic English captions to live streams whenever professionally provided captions aren’t available.

Download ‘Clips’ to understand why existing industries involving voice recognition should listen up. What it does is nothing short of amazing.

Just a few years ago no one cared about voice recognition. Alexa, Home and Siri managed to ship consumer products with voice as a central feature and we started to realise the potential.

We at Konch are in the center of this technological revolution and we believe that voice recognition will disrupt traditional industries as it becomes better and better.

In researching our market we interviewed various people that work as transcriptionists; specialized people that transcribe audio and video files for a living.

Most of the people we talk to either rejects the idea of computers disrupting the industry or believe it will take a long time. As some of the interviewees said: ‘Voice recognition is too unreliable’

In terms of voice recognition being too unreliable we are on the same page. Unreliability is what has kept voice recognition from becoming mass adopted.

We face several issues; regional accents and speech impediments can throw off word recognition platforms, and background noise can be difficult to penetrate. And simply recognising sounds isn’t enough — to have any level of effectiveness, systems need to be able to distinguish between homophones (words with the same pronunciation but different meanings) and learn new words and proper names.

We are not where we should be yet. So far so good.

However, the speed of innovation is accelerating and it’s not going to slow down. Currently, Baidu’s voice recognition is better than most humans at identifying spoken words with a 96% accuracy. Siri coming in second with 95% accuracy.

Baidu, Apple, Google and Amazon all have close to an unlimited amount of data to train their AI to become better. Where is that data coming from? Us. You and me. As an example, 65 percent of smartphone users reported using the voice assistants on their phones. Unlike you and me the deep learning algorithms are learning 24/7/365.

I understand why most transcriptionists see AI and voice recognition as a threat to their business. Not because it will take away demand but because it will redefine your role. I believe the job of a transcriptionist will become a shared responsibility between human and robot.

Finding that balance between machine and human is what we specialise at in Konch. We are not here to replace human transcriptionists. You are needed. Don’t leave. However, we believe the job can be done more efficient that benefits both the customer and the transcriptionists.

Remember, technology is coming. Embrace it and be open to adapt. Otherwise, you might end up like Blockbuster, Kodak, Nokia .. the list goes on…