Google's DeepMind is entering Skynet territory as it shared the news about cracking the code on the elusive Text-to-Speech system, which will make machines sound less like robots.

Skynet, of course, is the artificial intelligence system in a dystopian future in the movie "Terminator."

In a blog post, Google's DeepMind introduced the "WaveNet," which it said "mimics any human voice and which sounds more natural than the best existing Text-to-Speech systems, reducing the gap with human performance by over 50%."

The practical applications of this breakthrough are substantial. For example, those who lost their voice and depend on the machine to converse with others might be able to "find" their real voice again thanks to the WaveNet.

Imagine Stephen Hawking talking more like a human than a robot; or Siri for that matter.

Instead of the current system where the text-to-speech system relies on the human model speak random words and sentences, and the algorithm creates entire new words using stored data, the Google's DeepMind managed to model the "raw waveform of the audio signal."

According to the new article from Bloomberg, WaveNet is a neural network (much like Skynet) that impersonates some functions of our brains.

The bad news, however, is that Google's DeepMind still could not solve how to translate the WaveNet into commercially viable practical app. The data involved is just too big to compute.

But this piece of news by Google's DeepMind is nothing to sneeze at. In fact, companies are now closely looking at any developments considering the virtual assistant potentials of WaveNet, among others.