@Background Pony #7B16
Then they should have plenty of software staff through, again, the cash they gained from the advertisements to fund their HQ. YT has been going for 10 years now. Surely, they saved up a huge amount to get some crack shot crews for that type of thing.
@vevans0009
I didn’t say it was expensive. I said it’s difficult. Very, very difficult. Programming a computer to recognize one person’s voice is challenging at the best of times. Programming a computer to somehow recognize the general auditory patterns of speech and coherently translate those often very loose patterns into language is a goddamn nightmare. Just imagine how many times you, or other people you know, have misheard things said clearly. Or go watch youtube videos of people with particularly thick accents and see how often they say something that sounds very similar to a different word.
Then imagine that instead of being a human being who’s used to the language, you’re a parrot from russia who only knows english from watching The Golden Girls.
Closed captions are manually typed.
If local news stations can have perfect captioning then youtube should too from the advertisments. XD
The complexity of voice recognition is huge, and has been an open problem for decades. Solving problems isn’t always as simple as “throw money at it”.
Didn’t the name “Steven Magnet” originally come from the Youtube captions?
What about the bullets? Forgot about that. XD
If their crews are shooting crack, that explains a lot.
Then they should have plenty of software staff through, again, the cash they gained from the advertisements to fund their HQ. YT has been going for 10 years now. Surely, they saved up a huge amount to get some crack shot crews for that type of thing.
I didn’t say it was expensive. I said it’s difficult. Very, very difficult. Programming a computer to recognize one person’s voice is challenging at the best of times. Programming a computer to somehow recognize the general auditory patterns of speech and coherently translate those often very loose patterns into language is a goddamn nightmare. Just imagine how many times you, or other people you know, have misheard things said clearly. Or go watch youtube videos of people with particularly thick accents and see how often they say something that sounds very similar to a different word.
Then imagine that instead of being a human being who’s used to the language, you’re a parrot from russia who only knows english from watching The Golden Girls.
Naw, that can’t be it. They make a lot of money in advertisements so they would get the software easily.
Because good voice-recognition software is incredibly difficult.