So earlier this year, the company I work for lost some medical transcription clients because they decided to use AI. We have started getting them back, because it is taking clinicians too long to fix the AI transcripts. So now they pay for the AI, and then they pay us to fix it.
Interesting - we've started using AI for our notes - and I've been really impressed for the most part. It does, and always will require editing, but it is faster than me typing the notes. I'm still skeptical on if they are truly destroying the recordings after 2 weeks like they promised.
Natural Stupidity ?
Back in the day when I was a transcriptionist, this all started with my main client trying to use Dragon NaturallySpeaking. OF COURSE IT DIDN'T WORK. DragonSpeak didn't know the difference between perineal and peroneal, and that was just the start. Congrats on your new income stream. :D
Efficiency!
AI transcription is fine if you want to get the gist of something, let's say a celebrity interview. I could never imagine using it for MEDICAL purposes (nor legal, nor anything that could adversely impact a life).
Extremely interesting data point! Do they pay your company the same amount as before? Or is it possible there's still some net savings?
And the more they "fix" AI, the more it lies confidently, increasing the danger/risk to others. arstechnica.com/science/2024...
Human feedback to AIs makes them favor providing an answer, even a wrong one, while making the answer more convincing.
Tax on stupidity
It's Dragon Naturally Speaking all the way down.
My company is about to start using AI for psych and therapy notes. I'm... skeptical.