I found out about this thing where you can help with open-source voice recognition software. Partly that’s awesome in and of itself, but they’re also really hoping to get people of all ages and genders with as many accents as possible so that it works better and can understand people more reliably. They also encourage you to not find totally silent areas in which to work so that it can also differentiate the human voice from background noises and such (and we all know what a pain it can be to try and speak clearly, enunciate in a way that programs understand, and escape the ever-present background noise!).
It’s called Project Common Voice, and is being done by Mozilla. There are two different ways in which you can participate.
The first way is by reading given sentences in order to give them more data to work with. The second way is by validating *other* people’s submissions. You see the text they are supposed to be narrating and you listen to them say it. So long as they say all of the words correctly, you click to accept it as correct. If they say it incorrectly, you click the button to say so. I have found one person who says completely and totally wrong things, clearly on purpose, which irritates me, but for the most part it’s people honestly just trying to help.
I’ve heard many different accents, as well as people for whom English is clearly not their first language. Being a native English speaker, it never really occurred to me how difficult it must be for someone who isn’t fluent with the language and still has a strong accent to try and use voice recognition software. I’m really enjoying hearing the soft, uncertain voices speaking these phrases. I’m enjoying the feeling that I’m helping people, but also that I’m not the only one who cares.
Also, just for entertainment value: