So the conference of the Modern Languages Association, right? Emphasis on languages. As in many different languages. In my session, obviously, everybody is a Spanish-speaker.
Because Spanish is a language. Or so we thought.
At the beginning of my session, a representative of the association appears in my Zoom room and tells me that I have to turn on automatic speech recognition that will provide captions. For purposes of inclusion.
Sounds great, right?
That is, until you discover that this automatic speech recognition only understands one language. Which is THE language. English.
So what happens to those of us who speak other languages at the conference on modern languages? The speech recognition assumes we are speaking a defective version of English and transcribes all we say into English.
Mind you, it doesn’t translate. It transcribes.
I’m not allowed to turn off this garbage. As a result, it assigns to me things I never said and that the automatic speech recognition assumes I said. To my horror, I discovered that I spent half the session saying some sort of garbage about Taiwan. Which I never mentioned.
Another speaker was recorded as saying that “Oprah is a child controller.” She didn’t speak about Oprah or children.
Most shockingly, “Moshe should run Africa and go supreme” was a statement assigned to another speaker. I don’t think she knows because most people can’t follow subtitles as they speak.
The captions move so fast that it’s impossible to catch everything that goes into them.
At any other time, this would be funny. But the way things are, it’s downright dangerous to have a transcript associated with your name where statements are assigned to you without any input or control on your part.
This is how things always are with “inclusion.” Intentions are great but the result is both insulting and horrible.