If you ever meet iTnews for an interview, the resulting story will be written with the help of artificial intelligence.
I don’t mention this to talk shop, but to point out that if AI can change something as simple as an interview, it can change anything.
Here’s how it changed us.
We like to record and then transcribe interviews to make sure we quote people accurately. But transcribing audio is a tedious and thankless chore that requires about three minutes to turn one minute of audio into accurate text. Paid human transcription cost $1.50 a minute and takes forty-eight hours to turn around.
We can’t afford that time lag and the monetary cost is unwelcome.
So in late 2018 we started playing with AWS Transcribe and the Azure Speech-to-text API, cloud services that turns audio into text.
Our first experiments turned a one-hour interview into text in 30 minutes, at a cost of a single dollar.
While the cost and time were vast improvements, the transcriptions weren’t. The team figured we’d need to wait a year or three for the technology to mature.
It turns out we only had to wait until March 2019 when we found a smartphone app that delivers speech-to-text in real time and offers an online editor that makes it easy to search and review audio.
There’s nothing to learn, nothing to pay (for ten hours of transcriptions per month), no hardware to buy and no complications.
The app easily picked out deeply technical terms during a DevOps conference, a feat that won over the whole editorial team.
Our work practices have since changed. We finish stories faster and they include more quotes.
We think that’s because we can make more eye contact during interviews – an example of machines letting humans do the things humans do best.
This transcription app is not a secret. Our rivals have it and they’re changing too. So we’re now on the lookout for the next tool to improve our work, after years in which the way we conduct interviews has remained largely unchanged.
When the team started work on the 2020 iTnews Benchmark Awards, we wondered if this year’s entrants would tell similar stories of advanced technologies becoming affordable, then becoming ubiquitous, then rapidly changing established business processes.
We didn’t have to look far.
Many entrants see the ability to make better-informed, data-driven, decisions as a tool that will help them achieve market leadership.
Most recognise that building that capability is not a one-and-done proposition, and that they’ll need to become capable of rapid and constant change.
We’ll reveal the details of how entrants have put digital technology to work when we name finalists in the awards in January 2020.
Before then, iTnews has again teamed with Benchmark Awards partner KPMG to present a series of stories that explore the big issues facing businesses and technologists in 2020 and beyond – issues just like those that have empowered our team, and threatened others.
This year we’ve called that series “Digital & Disruption”.
Please don’t hesitate to let us know what you think of the series as it rolls out over coming weeks. If you’d like to do that during an interview and watch our team use AI transcription, all the better!
Simon Sharwood
Editorial Director
iTnews.com.au
