The Speech-to-Text Boxes are Coming

The Speech-to-Text Boxes are Coming

Roughly half of all online holiday sales in 2017 were done on Amazon.com, and the single top selling product was the Echo Dot, with millions of devices shipped. The Echo’s main competitor, Google Home, also sold over 6 million units since October of 2017.

When you strip these blockbuster holiday sales products down to it’s core, they’re boxes that were designed to capture natural language processes. Inside are advanced omnidirectional microphone arrays that can pick up what people are saying from any location in a room.

The beauty of the Echo and Google Home is that they represent the marrying of billions of dollars and decades of research and development of natural language processing. A major limitation of this immense undertaking was the microphone hardware required to capture the human voice with enough quality that the software can interpret voice wave patterns accurately. The Google Home and Echo represent a bridge of the hardware-software gap, catapulting us into a bold new speech-to-text revolution.

Echo’s are predominately used to turn on the lights, play a Pandora station, and set timers for our food. Soon they’ll be able to do so much more with AI neural networks rapidly advancing. What this means is they will able to understand much more complex sentence structures and then provide more complete actions and answers. We could go down the rabbit hole of possibilities, with images of the movie Terminator and Her coming to mind. The important point being this is a juncture in the way we interact with technology, with holiday sales numbers painting a compelling picture.

As Google, Microsoft, Amazon, IBM, and Apple race to evolve AI natural language processing, the legal and insurance industry that rely on humans to transcribe voice stand to see huge benefits.

Leave a Reply

Your email address will not be published. Required fields are marked *