A friend of mine works on developing speech recognition systems. The type of systems that Apple’s Siri uses in the backend. One of the major challenges such systems face – is interpreting the accents. We as humans sometimes find it hard to interpret some people’s accents, imagine how can software do it.
Software is no magic science. It is after all built by humans and for humans to make such software – they need to understand various accents too.
I have met people from lot of countries and everyone has their way of saying things – Turkish, Russians, Asian, South East Asian, British, Australian and so on. Sometimes, I find it really hard trying to interpret what the person is saying and I often wonder how this person is able to communicate with his team members. It must be really hard.
But I guess in technology world – it doesn’t matter what language you speak, all that matters is what language you know. And by language in the second half of my previous sentence – I mean programming language (Java, PHP,C++ etc). If you can program well, you get hired, doesn’t matter if you communicate well.
There is a person in my workplace who can’t pronounce anyone’s name properly. He will chew up everyone’s name even if it is a small easy name or a long hard name. For example – Richard becomes Richie, Anita becomes Ann, Sonia becomes Son and so on.
With time of course, you get used to the various accent that people around you use and start understanding them better. Also with time some of the younger generation changes their accents to match the local one. Software of course goes through various rounds of releases and bug fixes to ultimately understand all possible combination of accents.