Apple’s event this week brought a lot of interesting announcements. We saw updates with facial recognition software, a new Apple Watch and Apple TV, as well as wireless charging stations for all your Apple devices. However, we heard very little about a feature that usually is one of the headlining features each year: Siri.
Even with just minor announcements for Siri, it’s clear that virtual personal assistants are becoming a staple in electronic devices. High quality virtual personal assistants are becoming an expectation for consumers. No longer do people have to communicate through typing code or even clicking a mouse. The convenience of being able to use your voice shows a shift in the industry: gone are the days where humans need to adapt to the machine. The machines are starting to adapt to us.
Virtual Personal Assistants and Home Devices
Take for example smart home devices. 15 years ago, the concept of controlling any part of your home automatically seemed like a sci-fi, futuristic reality. Now, not only can you control parts of your home remotely, but today you can do it with the help of with your phone’s virtual personal assistant. Take a look at the “Home of the Future” from Microsoft in 1999:
However, virtual personal assistants today have their limitations. I’m not just talking about vocabulary and syntax limitations, which aren’t perfect by any stretch of the imagination, but about how effectively they fulfill the expectation of Assistant. We expect an assistant to know our preferences and be tailored to our specific needs and tastes. This is where we see Natural Language Generation (NLG) taking a role.
Benefits of Adding NLG to Personal Assistants
At the moment, virtual personal assistants fall under the umbrella of Natural Language Understanding (NLU), which focuses on how to help the computer understand human inputs, in either text or voice. Think of it as the computer turning our words into code or a language it can understand. You ask a question, the virtual personal assistant searches for an answer, and then reads back the results.
Natural Language Generation (NLG) technology focuses on the opposite effect, looking how to take a computer language and turn it into text that a human can understand. When these two technologies are coupled together, you take a step beyond reading search results. You can go from saying “Generate an analysis of Q2 results” and within seconds you can get a personalized first draft of the analysis.
Looking forward, voice assistants and similar technologies will only become more helpful as they can integrate seamlessly with a variety of technologies. They need to go beyond just being sophisticated search engine readers to something that can be more personalized and unique to the user. As we begin to interact more and more with computers in their variety of forms, being able to communicate and receive unique, yet accurate results will be crucial.