Perspectives

Do You Know Me? Can I Trust You? (Part 1)

By Eric Fensterstock


Personalization, Privacy

Um, my friend has a problem with his ____. Yeah, you know, my friend.

Yes, it’s an embarrassing problem. I’d better research it in my browser’s private mode. But Google may be able to figure out it’s me, so I’d better search with DuckDuckGo. I think they will protect me better. But won’t my ISP know what I search for? Should I use Tor to encrypt my internet activity? If I do, will the government think I am up to no good and start logging all of the goofy videos I watch?

Is it ok to search through my phone? Or is my carrier the biggest spy of all? Maybe I don’t care about privacy, but do I really want to see ads about this embarrassing medical condition for the rest of my life?

Personalization and the sharing of private data offer a wealth of benefits, but clumsiness and abuse threaten to undermine this future.

Everything is Awesome (or at Least Really Smart)

Smart homes, smart cars, and wearables are already here. The Internet of Things promises to embed intelligence in any object. The breakthroughs will come not just in the choice of things to be moved online, but in the ways different devices and sensors throughout our world work together. My sleep tracker will see that REM sleep has ended and start my coffee maker. Then my bedroom will gently light up as I hear that song I have been into lately.

Just by going through everyday life, each of us will produce a library’s worth of information. This data will be mined for patterns by constantly evolving artificial intelligence. If we consent to this analysis, the benefits will make targeted ads seem like primitive toys. A simple perk would be if Netflix recommended movies based on the books, music, and activities I enjoy, not just what I watch on Netflix. Spotify could let me know about the soundtracks to movies I have recently watched or the best bands that will be playing in the city I plan to visit. Recommendation engines can go further, analyzing me to suggest nearby restaurants, stores, and even potential friends. After I have a great restaurant meal, I’ll receive an easy recipe for the entree along with a store plan showing where to find the ingredients in my supermarket.

The Nest smart thermostat already saves homeowners money, and further home automation can increase savings. Our wellness can be improved by apps that monitor our activity, heart rate, and diet. Someday, health problems may be spotted in advance. Apps will be able to suggest healthy substitute ingredients for recipes and analyze restaurant menus to suggest the healthiest meal that agrees with the user’s tastes. Google Now already tells us when to leave in order to be on time for an appointment, and GPS has led to a generation that never needs to be lost.

Smartphones are on their way to near-universal adoption. According to BI Intelligence, there will be 24 billion objects on the internet of things in 2020 (vs. 10 billion traditional mobile devices). It will be hard to resist this new world. Embracing a world of data will mean being richer, smarter, and healthier.

When all of our data are considered together, society will benefit. Public health and urban planning will draw on massive real-time natural experiments. Traffic jams and accidents will be reduced by a coordinated fleet of self-driving cars. Academics of all stripes will gain insights from the analysis of big data, breaking new ground in social sciences and medicine.

There is a chance to create the next Google- or Facebook-sized behemoth, if Google and Facebook themselves don’t seize control of this opportunity. The winner will achieve unprecedented levels of customer intimacy along with intelligence into the constant movements of the world economy. Only a company that can execute with surgical accuracy and complete trustworthiness will capture this prize.

Big (Data) Headaches

The future is awesome and the future is terrifying. Just like today. We can already see the challenges that lie before us.

How are we going to lock down all of our most private data? Is this even feasible when our health data, home activity, and movement are being tracked? We don’t want data to fall into the hands of criminals or spammers. Even the well-intentioned can misuse our data.

Aggregate data can lead to smarter government, but we don’t want to live under the surveillance of a police state. We also don’t want our information released to the general public. The world doesn’t need to know every time the smart toilet flushes.

Beyond security, data analysis holds pitfalls. How many times have you seen product recommendations on ecommerce sites that were laughably wrong? This is the “TiVo think’s I’m Gay” issue, which was a source of laughs and consternation at the start of this century. If automated systems do more than just provide recommendations, if they act on their conclusions, all kinds of mischief are possible. Imagine “wacky misunderstandings” of the Jack Tripper and Lucy Ricardo variety. The worst case? War Games and 2001 A Space Odyssey.

One solution is to allow user feedback. Google at times asks users whether search results are relevant, and Facebook allows users to evaluate ads and posts in newsfeeds. Users of the Amazon Echo device can provide feedback on the Alexa system’s voice recognition. Virtual assistants like Apple’s Siri that may not explicitly ask for feedback will need to pick up on the hints and social cues that humans naturally learn to read.

The Starting Pistol Barely Fired

This post just discusses a few of the hurdles on the way to more intimate computing. In part 2, we examine further hazards and offer some solutions. The race has just started. As we pass each milestone, we will see a better world, as long as we don’t trip ourselves up.