How artificial intelligence is helping you get closer to owning your dream droid

Artificial intelligence has come a long way, and the days of owning our own R2-D2-like robots are not far into the future. Why? The necessary technology that can actually understand human language, interpret it and answer accordingly, is already on its way.

Are you excited about the new Star Wars movie? I know I am. My favorite characters are all the droids. Every time I watch the movies I can’t help but imagine how amazing it would be to have my own R2-D2. The great news is, it’s not as far-fetched as one might think. Technology that can enable conversational and smart robots is more real than you may know.

Think about it: Siri, Alexa, Google Assistant – all are body-less assistants that help answer your questions, find information you are asking for and crack a joke every now and then. And to be honest, even though the beep – beep – boop of the Star Wars droids is funny, I like the real assistants better as I can actually understand what they are saying.

Today’s technology is already capable of the ability to understand a human, interpret what was said and answer accordingly. It is amazing how far we’ve come. Years ago, to communicate with a machine, we had to develop code in machine languages like Assembler. Since the invention of voice recognition we are capable of converting speech into something the computer can work with and now natural language understanding enables machines to put the individual words into context.

But we are not there yet. Ken Arakelian, Director in Nuance’s Cognitive Innovation Group, just recently talked about it in his blog post “What’s left to tackle in voice technology – Part I”.  As you can read in his article, understanding language is more than just looking at individual words. It is all about the context, and sometimes context that was not even mentioned in the request. If you ask a virtual assistant, “Is it raining in New York?” it may be able to answer “yes”- but when you follow up with, “What about Boston?” it is left confused, unable to understand you are asking if it is also raining in Boston.

For a conversational experience, especially when the virtual assistant is designed to solve problems, the intelligence has to look at everything that was said and pull information from historic or 3rd party data to understand the bigger picture and  make sense out of it all. The good news? This exact technology is being developed today and soon machines of all kinds will have virtual assistants that are able to intelligently communicate with humans.

So while my dream of having a rolling friend just like R2-D2 may be a few years off, we’re accelerating machine learning to a point where it’s not so far-fetched. For now, I will continue dreaming about  jumping with it onto an X-wing Starfighter and saving the galaxy.

Customer engagement powered by Artificial Intelligence

Do you want to learn more about Nuance AI capabilities? Click here for more information.

Learn more

Tags: , , , , , ,

  • Div

    Artificial intelligence is the simulation of human intelligence by machines.nice one

Josefine Fouarge

About Josefine Fouarge

Josefine Fouarge is Sr. Product Marketing Manager in the Nuance Enterprise Division, focusing on automated and human assisted engagements in digital channels. Josefine brings more than 10 years of experience in sales and marketing for technology related businesses in Germany and the U.S. Her past expertise ranges from selling and configuring Apple computers to defining the market and messaging for a security software for on-premise datacenters and services offered through the cloud.