This is the first Ezine article that I am writing in about 15 years, and looking back, the world has changed so dramatically – generally, and my own, that my outlook, priorities and couldn’t be more different than I could ever even have imagined.
Nearly 14 years ago I stumbled upon online freelancing and by default became a transcriber. Long story short, I entered the world of television and film, having first transcribed focus groups, telephone calls, clandestine interviews in noisy restaurants, and more ghastly stuff than I care to remember.
My entry into film and television changed everything, starting as a lowly transcriber, climbing my way to running one of the most successful television and film transcription businesses in the UK. I don’t own it, I just run it. Just running it means orchestrating approximately 70 transcribers and around 500 clients. Within a year and a half I admitted defeat with serious burnout and had to cut down to half days only, giving me a great deal of free time to explore the changes that have happened in the past 15 years.
And I stumbled upon speech-to-text AIs, which are revolutionising the world of transcription.
Now, who needs transcription services?
And what services do transcribers provide for them?
Post production scripts
Raw interview transcripts
Interrogations for prosecution purposes
The list is endless.
When I first found out about speech-to-text, it was when I purchased Dragon Naturally Speaking, ooh, in about 2014. I found myself taking longer to edit the transcripts than if I’d just transcribed them from scratch.
Enter Amazon Web Services.
Amazon Web Services has created the most powerful speech-to-text AI that the world has ever seen or is likely to ever see.
Thanks to devteam.space, I came across the following statistics:
Markets and Markets did a study that concluded that the speech to voice recognition market would go from $7.5 billion in 2018 to $21.5 billion by 2024 with an annual growth rate of 19.18 percent.
So, I jumped at the opportunity and it’s taken me around eight months so far to get my own speech-to-text AI developed from the ground up. I’ve been scammed, conned, lied to and cheated by unscrupulous developers and have almost been financially ruined by pressing on, in a bid to gain access to this industry.
And in about two weeks from now, my AI is going to be ready for the world to use in any which way it seems fit, to create the most affordable, high quality transcripts that any AI can offer. Welcome to accuscript.ai. At a mere $0.075 per audio minute and an average transcription time of 10 minutes per audio hour, with 90 percent accuracy (and improving), this is a win-win for everyone – AI owners and users alike.
No longer do we have budgets that allow us to use human transcribers starting at around $1.20 per minute!
And don’t be fooled, those transcripts have been put through an AI first, and just lightly edited by a human being. Imagine that – $1.125 per audio minute for editing!
Yes, I worry about the transcribers (myself included) who rely on the high rate charged for human transcription. But we have to stay ahead of the time, or at least try to keep up with it.