Yo, Alexa!

My first experience with voice technology was listening to my dad make calls using bluetooth to connect to the speakers in his 2005 Acura MDX. I remember being impressed with the technology and specifically the fact that a machine was able to interpret our language. Removing the need for hand controlled devices makes driving much more efficient, and safe. The car came with a free, year long subscription to a voice application called On Star. On Star was advertised as having these capabilities:

1. Automatic Crash Response
2. Emergency Services
3. Crisis Assistance
4. Roadside Assistance
5. Navigation

All of these features were accessed by a small button behind the steering wheel that turned on a microphone, and from there my dad would engage with the app through voice prompts. It was extremely useful for him at the time because it was convenient. He was able to drive hands free and get turn by turn navigation to where he wanted to go. It was an example of a successful integration of technology and the real world. At that time the On Star application running in my dad’s car only knew a few a things about him.

• his location
• his phone number and contacts
• his voice, if the microphone was on

That was over ten years ago. Since then technology has come a long way. Now, technology knows more information about us than most of our friends and family. Most smart phones know an extensive amount about us.

• our location at all times, not just while we are driving.
• our financial information
• our identity
  • thumb prints
  • passwords
  • facial structure
• our heart rate
• our interests
• our conversations
• weird journeys into the internet

These are just some examples that voice devices like the Amazon Echo Dot, Google Home, all have access to.

As a side note I am fascinated by what this will mean for society as these technologies improve and become more integrated in our lives. Specifically, I wonder how talking to a infinitely smart machine who knows us well change society.

I think the adaptation of voice technology and smart phones are similar. Both represent a drastic change in the way we think, act, and feel. This technology is going to advance exponentially, and I am specifically interested how interacting with essentially infinite intelligence with have on our thoughts, feelings, and emotions. Will it be helpful or harmful? Here’s the way I see it.

Helpful

• it removes a step (the scree) making it more efficient.

Harmful

• if people become over reliant on voice tech, how will that impact other skills such as person to person communication or writing?

Reading about the way voice interactions were designed gave me an idea. A fundamental aspect of the way Alexa was programmed was to make her seem as human as possible in her responses. For example, if you’re having trouble hearing someone you might wait and try to pick up what they mean instead of interrupting them immediately. Alexa does that too.

The largest challenge that voice technology has to overcome will be privacy and capitalism. In order for Alexa to work effectively she has to be listening to you at all times. We don’t sacrifice that amount of privacy even to those closest to us, yet we are supposed to give permission to Amazon?

Capitalism is a threat because they own the increasing amount of personal information that devices mine from our lives. Will they use it for good, or evil? To answer that we can look what is already happening. It’s clear now that every moved you make is tracked, sold, and then used to try to sell you things. Facebook has taken that a step further by reportedly spying on conversations and creating targeted ads around that.

It was recently reported that Google targets ads to users based on email content, and despite outrage, still read the content of our emails.

There are solutions to this, and its all based on encryption. Apps like Snapchat and WhatsApp implement encryption into their services that deletes information they’ve gathered after a certain amount of time. But will people trust companies to do this? Snowden wouldn’t. And it’s very difficult to test.

A huge issue with Alexa now is her reliance on the cloud. Every interaction with the technology has to be relayed to Amazon’s cloud through their servers, and back. That is where the danger lies in protecting privacy.

Last year MIT released a chip made especially for AI. Essentially their goal is to make a chip that can handle the immense computing power that AI requires locally on a device, so that it removes dependence on external clouds and services. This could be another way to make privacy concerns decrease.

Developing this type of technology will be heavy. There are lots of ethical questions to weigh while designing a device that has so much access to sensitive information. Things have come a long way in ten years. I’m excited to see what things will be like in another 10. Hopefully I won’t be longing to go back to the days when the voice technology was as simple as making calls for my dad on the way home from his weekly basketball games.

Building An Alexa Skill

I borrowed an Echo Dot from school and watched this tutorial on how to build a skill. I was pleasantly surprised on how easy Amazon has made it to learn and build skills. Here is what I did:

• Activate my Echo Dot by using the Alexa app on my phone
• Make an Amazon Developer Account

• Created a new custom intent called MattIsolaIntent within a skill called Critical Making, which was created by my professor for our class. This graphic helped me visualize the process.

  

• Created some utterances that activated my intent

  

• Tested it

  Alexa, what's my name? from Matt Isola on Vimeo.

What I learned

Voice technology is the next big thing. This type of interface opens up a whole new world of possibility for innovation.