Google has a dream that is akin to what some science fiction movies have wherein artificial intelligence within those flicks are so smart that they can even create witty retorts. Just recently, the Internet giant has just reached a milestone with its DeepMind artificial intelligence project as the successful development of the technology can now mimic the sound and way of speaking of the human voice.
Google’s DeepMind AI is the Company’s Latest Milestone
The new technology within the Google AI is called WaveNet, and it was a breakthrough that is described as a deep neural network in which it can generate raw audio wave forms in order to create speech patterns. This is reportedly to beat the existing Text-to-Speech systems, even those coming from the search engine firm itself.
As per researchers within the Britain-based WaveNet team, the gap in human performance has now been reduced by as much as 50-percent. In their studies, human conversation can be demonstrated in an actual AI. It might be also interesting to know that the WaveNet technology being used us capable of learning different types of voices and speech patterns. So much so that it can even simulate mouth movements and artificial breaths. Furthermore, it can even emulate emotions, accents, and language inflections.
“A single WaveNet can capture the characteristics of many different speakers with equal fidelity, and can switch between them by conditioning on the speaker identity,” the researchers wrote in a study. At the time of writing, the WaveNet technology by Google is capable of using the English and Chinese languages. Also, it can even produce music such as classical piano pieces, and it can even compose songs all on its own.
What is the importance of this breakthrough by Google in artificial intelligence technology? This is a cause for celebration for the tech world, especially for the Internet giant. This is because the AI rests on the sheer amount of data that is required to achieve its current quality. In other words, you should consider how many computer-generated systems are using Text-to-Speech technologies that have the same notion when it comes to gathering data. By pooling in what has been gathered into an artificial intelligence that knows how to talk like a human being, and in an emphatic way even, then we might be one step closer to the AI that can be seen in those science fiction movies and TV shows.
Share This on Facebook