Talk to Your Device
In the original Star Trek TV series, the crew of the Star Ship Enterprise would simply say "Computer," and the ship's computer would respond in a computer voice with an answer to their questions. Fifty years later, voice response capability is nearly ubiquitous working with devices small enough to fit into our pockets. Today, you can talk with your phone, your TV remote, your computer, or though special devices like the Amazon Echo or Google Assistant.
In general, voice response comes in two flavors, text to speech and voice assistant. Text to speech has been available on computers since the 1980's with products like Dragon's Speaking Naturally product. Most recent Android phones provide the ability to use your voice instead of the keypad to enter text into messages and most other products. The voice recognition works with the keypad application so it works consistently with most Android applications that use the keyboard. Some early phones had a simple voice command capability that understood a few simple commands like dial home, but these could not handle continuous speech like modern phones.
Apple kicked off the current age of voice assistants when it released Siri. Siri did many of the same things as was available in Google Voice Search but did it in a more conversational format. In addition to basic search functions, Siri could answer questions posed in normal conversational English, and Siri could tell jokes. To access Siri on a Siri-enabled device, you simply say "Siri" followed by what you want to ask Siri, and Siri will try to answer.
Google Now to Google Assist
When Siri was release, there were a lot of comparisons between Siri and Google Now, which followed Google Voice Search. Google added support for a number of Google app and phone commands. You could simply say "Ok Google, Navigate Home," and the device would launch Google Maps and create a route to guide you back home. Google Now is being replaced by Google Assistant. Google Assistant is supposed to be more conversational than the "Now" version and is backed by an Artificial Intelligence (AI) engine. Google collects a lot of information about your online activities and tries to apply this knowledge to return the best answers that are relevant to you. Google now offers a voice appliance called the Google Assistant. This is basically a smart speaker with a built-in microphone. This concept was pioneered by Amazon.
While Google knows what you are doing on the Internet, Amazon knows what you buy. Amazon created the Echo and gave the Echo the Alexa voice assistant. The Echo is basically a smart speaker with built-in microphones and no screen. With Alexa you can order things from Amazon, and they will be automatically sent to your home. This has had some interesting consequences with children ordering toys and demonstrations on TV causing many orders from Echo devices hearing and reacting to the TV orders.
Amazon not only gave Alexa the ability to order things via voice; it created an open programming interface and encouraged other companies to create voice applications. Amazon also enabled the machine to connect to Internet of Things devices you might have in your house. This allows you to say "Alexa, turn on the lights," and if you have supported smart lights like the Philips Hue, the lights will turn on.
It is sometimes easy to forget that Microsoft creates Windows for both computers and telephones. With both Android and Apple providing voice capabilities on their phones, Microsoft needed to do so as well. With Windows 10, the Cortana voice assistant arrived on our laptops and desktops. If you have a microphone as most laptops do you can enable Cortana to respond to your voice by saying "Hey, Cortana." Cortana can search for applications and files on your PC, use Bing to search the web, and use Bing maps for navigation. It also integrates with several Microsoft apps available in the Windows app store.
Many of the current voice assistants now have the ability to translate words spoken in one language to another language. Google has been doing this the longest with its translate application, but most of the others support this feature now. To do this on Google, simply say "OK Google, Translate 'Hello, how are you,' into Spanish," and Google will respond in voice "Hola, como estas."
How Do They Work
In general, the devices we are discussing are not powerful enough to accurately understand every word we say as continuous speech. So while we are talking, the device captures the sound and sends it over the Internet where more powerful computers translate the voice into something the machine understands. The translated speech may be sent to another server to do an Internet search for example, or it may be returned back to the device to drive a command on the device.
Does this mean the device is always listening and sending what it hears to the Internet? Isn't there a risk that they hear everything going on around them? For the devices discussed here, the answer is no. These devices use what are called "guard words." The device itself has the ability to understand a few phrases without needing to send out to the Internet. When you say "Ok, Google" or "Hello, Alexa," the device itself recognizes the phrase and sends what follows to the servers on the Internet to process. When the interaction is over, the device stops sending what it hears and only listens for the guard words.
That is not true for all devices. It was demonstrated a couple of years ago that certain Samsung and other voice command televisions were listening to everything going on in the household and sending the information to a third party for processing. Essentially these devices and companies had the ability to eavesdrop on everything going on in the home. It was revealed in the news a few weeks ago that our government was aware of these weaknesses and leveraged them as they felt was needed.
Enough for now: enjoy the heat and have a heart-to-heart conversation with your phone.