Alexa will now respond in the computer-generated voice of Bachchan named “Alexa Bachchan”.
One who cannot recognize Bachchan’s voice must be from another world. Isn’t it amazing when Alexa will give a response to your “How’s the weather, Alexa” with the stunning voice of Mr. Amitabh Bachchan?
Alexa is programmed with the baritone of Mr. Amitabh Bachchan. You can now enjoy this superstar voice asking for your order. Amazon integrated the computer-generated voice of Bachchan with Alexa features. The company stated that it is not a recorded voice.
This stunning technology is the result of artificial intelligence and audio technology which converts natural language into machine language and gives response like a human brain. Human brains and computers work in a different pattern. A computer has the capability of crunching a long bunch of numbers in no time. But it cannot understand simple natural language like “Open the window”.
It needs artificial intelligence to understand such natural language where each person has a unique pattern. The technology recognizes speech patterns and compresses it in different ways. There are grammatical errors in spoken language but the system has to understand its context with intelligence.
The system has to understand grammatical errors and errors in spoken language. Humans may speak with a metaphor like “he is a night owl”. Here sentence does not mean that the person is an owl but refers to his habit. Sometimes human gulp words in the sentence that are meant to be understood by default. Like when a person says “meet at six”, it is for the 6 PM, not 6 AM. The natural language also has repetitions, pauses, and hesitations. Human voice may sound sliced or broken too.
With all such challenges of spoken language audio technology also have to understand multiple sound interruptions. Each of the alphabets has different phonic sounds. The same word can be pronounced in different ways with the phonemes difference.
Humans have variations in their accent according to the language and region. The same phoneme may sound different for each human. It is nearly impossible to learn all these patterns and recite the whole dictionary for a person. This goes with the Bachchan too.
Modulation can add excitement and fun the Bachchan’s voice. It may otherwise feel flat and dull like a bad dubbing film. Alexa and Siri do not record each word. Funny thing is Alexa can now speak languages that are not known to Bachchan.
The voice-operated assistant, Alexa needs training regarding the specific owner and their accent. This is a contextual learning pattern which may take hours. It is a hard process for a system to understand what the user wants to say.
Alexa allows third party skills which can be integrated with its existing program. This is a business sense obviously and it is predicted to be successful. India has more than five hundred million users and unlimited data packs at a cheap cost. These aspects will surely be helpful to this new voice addition in Alexa.
Leave a Reply