Once a subject of science fiction, virtual assistants are on the rise, offering more proof that we live in the future.

Virtual assistants are artificially intelligent computer systems (like Siri, Amazon Alexa or Google Assistant) that interact with users in spoken language. In 2017, there were over 70 million Americans using virtual assistants. Here's how they work:

A virtual assistant is a cloud-based service that can perform tasks and answer questions through conversational speech. Examples are Siri, Amazon Alexa, Google Assistant, and Microsoft Cortana.

Virtual assistants are not to be confused with smart speakers. For example, Amazon Echo and Google Home are smart speakers, or the devices used to access a virtual assistant.

During everyday use, the task of a virtual assistant is very simple — you ask it a question, and it answers the question.

In order to do this seemingly simple task, the virtual assistant must understand the words that you’re saying, make sense of the words, identify the response, and then deliver the answer in a way that you can understand.

What makes this type of technology possible is an interconnected solution of powerful computers, machine learning, artificial neural networks, and really advanced math.

To further explain, artificial neural networks use math to imitate the way brains function. It’s a system of neurons (which are really just numbers) that are interconnected into a complex network. The network then trains itself using artificial intelligence techniques to refine its accuracy about the input’s meaning and about the user’s desired result.

Here’s the step-by-step process that happens when using a virtual assistant.

Step 1. A "wake word" initiates a task.

Depending on your device, this can be prompts like “OK, Google,” “Hey, Siri,” or “Alexa.” The device uses a low-power neural net to constantly listen for these prompts that indicate the user is about to initiate a task.

Step 2. Translates speech to text command.

After hearing the wake word, the virtual assistant initiates a task by listening to your command and then transmitting the recorded question to a server. This data center houses a more powerful neural network which can turn the main question into text.

Step 3. Uses natural language processing to understand the task

The intended meaning of the command must then be deciphered. This is done via machine learning and natural language processing.

Step 4. Takes action.

The assistant then takes action on your command. It can use either a company’s internal information, your accounts, or third party sources to complete tasks such as an Internet search, playing music, getting directions, completing an online purchase, or delivering a weather report.

Step 5. Talking back.

In this last step, the device delivers its response to you. Again, machine learning is one of the primary technologies that are used to generate text-based responses and then synthesize them into natural-sounding replies.

Cool, huh? But smart assistants aren't without their risks.

It's true that smart assistants with enabled wake words are constantly listening. But they don't record your voice until they recognize their wake word. Once awoken, they do record and remotely store your voice speaking to the device.

Not only can these devices be a concern for data security, but they can also have legal repercussions. For example, in one Arkansas murder case, Amazon was given permission by the defendant to provide audio from an Echo smart speaker as evidence.

For most digital consumers, having online activity tracked is nothing new. But smart speakers are different because they track not just what you type, but also what you say in your home.

Just a few years ago, consumers were shocked by Samsung’s recommendation for users to watch what they say around their smart TV. With the data capturing abilities of virtual assistants, it’s worth considering the privacy implications before you use a smart assistant.

Did You Know?

Your virtual assistant can do more than just perform commands, it can also provide you with useful information such as your assistant command history.

To access your history on a Google Assistant, open the Google Assistant app. Then select Settings, and then Accounts & Privacy. Now select My Activity.

To access your history on Amazon Alexa, open the Alexa app, and then select Settings. Now select General, and then History.

There are some fun and surprising applications for virtual assistants that you may not know about. While some virtual assistant commands are practical and useful, other commands are insider jokes that let you have a little fun with your virtual assistant.

Try some of these commands:

For Amazon Alexa:


  • What should I wear today?"
  • Read me a recipe for buckwheat pancakes."
  • Order a large pepperoni pizza from Dominoes."
  • Buy more twizzlers."
  • Let's work out."

Try these Alexa Easter Eggs:

  • "Tea, Earl Grey, hot."
  • "Beatbox for me."
  • "Sing me a song."

For Google Assistant:

"Okay, Google..."

  • "What do I have scheduled today?"
  • "Play some 80s music.”
  • "Add raspberries to my shopping list."
  • "Play House of Cards on Netflix"
  • (From your phone) "Broadcast: ‘Hey honey, I’m headed home.’"

Try these Google Easter Eggs:

  • "Self destruct."
  • "I'm feeling lucky."
  • "What does one human sound like?"

For Any Virtual Assistant:

  • Set home security system
  • Add events to your calendar
  • Control your smart thermostat
  • Check your bank account