Web
Analytics
  • Digital Life Initiative

DL Seminar | Seductive Surveillance and Social Change: The Rise of the Voice Intelligence Industry


Individual reflections by Shirley Matcha, and Yinsongti Tindana (scroll down).



By Shirley Matcha

Cornell Tech


It probably comes as no surprise when I tell you that your phone is listening to you. Far too many have experienced odd “coincidences” that make us suspect our laptops or other smart devices have been eavesdropping on our private conversations and using this information against us. From my experience, I started noticing product advertisements in Spanish ever since I downloaded Duolingo to raise my Spanish-speaking skills from mediocre to slightly less mediocre.


And I’m sad to say that, as much as we all hate companies violating our privacy and exploiting our data to manipulate us into buying products we do not need, things will get a whole lot worse if we choose to stay silent about the emerging Voice Intelligence Industry.

Coined by Professor Joseph Turow of the Annenberg School of Communication at the University of Pennsylvania, the Voice Intelligence Industry encompasses the use of smartphones, smart speakers (Amazon Echo, Google Home), car information systems, customer service calls, etc. to collect information about a person’s voice. Now in the previous example, my phone was picking up on the words and overall content of my speech. According to Professor Turow, phones are mining for more: patterns of speech and the physiology of a human voice, like pitch, volume, tone, nasality, just to name a few. And although this news may sound a bit underwhelming at first, bear in mind that there have been decades worth of studies demonstrating that assessing these vocal features can allow one to accurately predict an individual’s physical, mental, or emotional features. This includes characteristics like weight, height, age, ethnicity, and even contraceptive use. We edge closer to living in a world where smart devices can racially profile people or car systems can detect early signs of depression, all based on the sound of one’s voice. We already live in a world where automated phone service bots can detect when a caller is angry, look at the caller’s customer history to determine if they are valuable to the business, and then use this data to offer them a special discount (this offer is presumably not given to those who sound calmer over the phone). This kind of discrimination is part of the reason why many like Professor Turow call for the outlawing of voice biometrics analysis altogether.

Yet, this technology has already found its way into our lives, and it will continue to integrate itself and become more commonplace absent resistance. Some companies are actively trying to get people used to voice-activated technology, as exemplified by Amazon’s “Alexa Evangelists” who tout the benefits of an encroaching voice intelligence sector. Additionally, Amazon purposely slashes prices on commercial holidays like Amazon Prime day to get people to buy devices with voice assistant technology. Other companies are more discreet. Instead of trying to convince you to opt-in, they make the choice for you. For example, BottleRocket Wine & Spirit has decided to replace some human store clerks with Amazon’s Alexa, who can ask customers about their alcohol preferences and then offer suggestions within the company’s catalog. Sneaky tactics also include burying grants of consent in the terms of service; companies allow you to use their smart devices in exchange for the permission to use your voice however they please.

When questioned as to why these companies are keen on using your voice, an Amazon representative stated, “This is a public service we are providing.” Put in context, Amazon was granted a patent in recent years for a method by which Amazon Echo’s voice assistant, Alexa, could hear a person’s voice, perceive if they are ill, and suggest purchasing food or medication. Google’s answer is less reassuring: “Well, we aren’t using this information for marketing now.” Note that Google has collected vast amounts of data about its users, through searches, purchases, and social media usage. So much so that some, like Eric Schmidt, predict that Google may be able to tell us what jobs we want to pursue, knowing more about our own selves than we do. Companies try to persuade us that they are stealing our information to make us better people or outright deny that they are using our information for their benefit, but the clear motive is profit. Besides marketing, the other looming threat is the potential for imposters to pose as you.

Voice authentication is the not the far future, it is the here and now. In prisons, computer programs are listening to prisoners’ voices when they make phone calls to verify the identities of prisoners and check the prison database to ensure that those individuals are allowed to use the phones. Professor Turow does concede that the use of voice recognition is a benefit when it comes to identity authentication, protecting one’s information or preventing fraud, since voices are fairly unique to each person, similar to facial recognition. However, it is increasingly obvious that the downsides of this technology will outweigh the upsides. The powerful combination of using an individual’s face and voice has further bolstered the already deeply troubling phenomenon of “Deepfake” videos by making it more difficult for people to see the difference between reality and edited content.

The disadvantages of adopting voice intelligence technology are being downplayed by those who want us to continue to remain passive until we are so enmeshed that there is no reasonable way to escape it without forfeiting certain advantages of technology. Professor Turow and similar advocates have sounded the alarm on the pressing issue. Whether we decide to heed his warning will invariably change the future of how we use technology and how technology uses us.


By Yinsongti Tindana

Cornell Tech


In 2003, a new show premiered on national television, first of its kind I had seen on our little Sony screen in Tema, Ghana. “Big Brother Africa”. It was a reality show where an invisible figure “Big Brother” was always watching the participants. Some years later, I came across the book 1984 which covered themes of mass surveillance and presented to me, a new kind of “Big Brother”. Whether for entertainment or for political gain, the thought of surveillance has always been a pretty uncomfortable idea.


For the second session of the Digital Life Seminar for the 2020/21 academic year, Prof. Joseph Turrow of the Anneberg School of Communication at the University o Pennsylvania explored this uncomfortable idea as he led a conversation on the complexities and concerns of seductive surveillance and social change, in particular, taking into account the voice intelligence industry.


What exactly is the voice intelligence industry?


Prof Turrow started off describes this emerging sector as involving the mining of voices through smartphones more speakers. Researchers have mapped relationships between the sounds of person's voice and features of that individual. Research argues that voice can be related to weight, height, age, ethnicity illness contraceptive use by the way you can you can tell a person's contraceptive views within the month of that person's taking a contraceptive by the person's voice and more.


Personalization is the modern-day marketing currency. When marketers apply these technologies to potential customers hyper personalized inferences about individuals can lead the marketers to perform social typing and provide them with differentiated opportunities, based on profiles, driven by machine learning and predictive analytics. This according to Prof. Turrow is significantly more attractive than traditional forms of personalization in marketing and advertising because people lie in the hope of deceiving marketers.


Why is the voice intelligence industry a source of concern?


These new technologies are able to detect the mood of a speaker t the sound of their voice and make suggestions to lift those moods. For example, a simple

“hey Google” with a course voice could get you a chicken noodle soup recipe recommendation or a suggestion to order some cough drops.


While this may sound like exciting technological advances, it comes along with some serious issues to consider. Prof Turrow indicated that this advancement signals a new era for discrimination and opens the floodgates of voice profiling in all sectors of our lives.

Companies have swayed populations to use technologies for the surveillance aims of those companies, consequently turning new digital developments involving surveillance into habits of everyday life, and then make uncoupling from them extremely difficult.


It is hard for privacy law to stop this because the direction of personalization. In general, as well as in specific advertising seems to be towards permission. The notion is that if you get permission from people. You can do virtually anything.


The way forward.


Prof Turrow suggests the need to promote perspectives and policies to do rail the technology. The seductive actors always play down the surveillance bit and market all the exciting parts. As such, consumers ought also to be on the lookout for privacy policies that allow entities use their voice data in any manner they deem fit.


The entire seminar was very eye opening and opened up a conversation about the need to protect privacy, vis-à-vis the wonderful convenience such technology provides. The convenience of modern technology is incredible but the balance between technology and privacy is difficult to create and my give rise to abuse by businesses.

Contact

Address

Cornell Tech

2 W Loop Rd,

New York, NY 10044

Get Here >

DLI Queries

Jessie G. Taft

jgt43@cornell.edu