Guest Blog | What’s Next with Speech Recognition? – Nuance

Everyone thinks that speech recognition is what they have on their smart phone.

But speech recognition is much bigger than that, and includes solutions for document productivity, intelligent assistants, command and control, and voice biometrics. These have application in business, healthcare, customer care, cars, and mobile solutions, to name a few. Nuance is a leader these areas, as seen by formal awards and product reviews.

Nuance R&D is helping to create these new technologies.

In Nuance’s traditional Dragon Windows and Mac speech recognition products for professional document productivity, the latest version 15 software uses machine learning to improve recognition of syllables and words, and improves performance in noisy environments. Surprisingly perhaps, machine learning uses fewer computing resources with the machine learning implementation.

In other areas, AI and machine learning are being applied to the interaction between people and technology. Nuance’s Nina is a virtual assistant that allows customers to transact, navigate, ask questions and get exactly what they need whether they are using a mobile phone, static web pages, or a mobile app.

Nina’s sister, Jetstar Jess, is Australia’s most popular airline employee “The Ask Jess program has really gone way beyond our expectations … About a year-and-a-half ago we had real problems where customers could be waiting 45 minutes to an hour to get through to our call centre, now most of the time we have an average answer time of a couple of minutes.”

Here, AI is used to add context to a conversation that might start as a conversation on a phone, move to the web, and then finish as voice running on a mobile app on a TV in the customer’s lounge room!

The question is raised about these new systems taking over the planet and perhaps deciding to terminate the human race. But we’re building artificially intelligent systems, not artificial humans.

I don’t see any reason in principle why humans can’t eventually build systems with human characteristics like self-awareness.  But while we have close to enough computing power to build something with the performance of a human brain, we lack the understanding to build a sentient being. We don’t know how to do it and it’s a problem that no amount of Big Data will solve. A solution is a long way away.

In the meantime, speech recognition and AI is helping us all be more productive.

The Nuance team, including Derek Austin, will be attending JuiceIT – 2017 Beyond the Hype!

Click here to learn more about the event and to register.