The Computer Just Found a Voice – and it’s almost too good

Deepmind (Google’s AI team) has created a super life-like Text-To-Speech (TTS) AI, that is very close to completely natural.

There is nothing more human than a conversation - hearing the spoken word, with all its cadence, ripples, cracks, and emotion. As a headhunter, my voice is my number 1 tool. In fact, as a human it’s my voice is my number 1 tool for so much expression.

The application of this technology is wide ranging and exciting. Here some potential applications:

If you’re a massive fan of audio books – like me – then imagine choosing between 5 different performance styles, in any accent, in any language.

The potential this tech has for talent management is also significant because it's the conversation that makes all the difference. The AI could assess interest, provide further information, and maintain contact with an authentic human accent. Couple that with the reliability that computers offer - and now we have something compelling.

This is no longer blue sky thinking - it's clearly within reach - within our lifetime.

Personally, I lean towards the positive potential of humanized AI. But automation of human work is a double-edged sword. Its negative potential should be considered less we face a backlash not dissimilar to the issue of economic migration, and globalization.

I strongly believe good work gives us purpose and leads to happiness - so we should take care to consider how we use AI to enhance our society.

Here is the article by Deepmind. Half way down, there are speech samples. Have a listen, and let me know what you think…

https://deepmind.com/blog/wavenet-generative-model-raw-audio/

William Laitinen – Exige International