Your TV Knows Your Dating Pitch
New ways of dating, TVs that spy, and a computer can see
...Raya is less superficial than its reputation would suggest. But surely Mr. Gendelman would admit that attractiveness and social capital play some role in its criteria for admission. Consider, I said, my nonexistent Uncle Tony — a hypothetical terrible-looking old man with no public profile and no Instagram following to speak of. If he applied to Raya, he’d be an automatic rejection, right?
Mr. Gendelman shook his head.
'Is Tony decently interesting? Is he passionate?' he asked. 'I’m not kidding — we’re interested in curating digital dinner parties, so to speak, and that comes in all forms.' [Kevin Roose, New York Times]
Don’t Swipe Right At All
For many, love doesn't occur at first sight. It weaves through a laughable maze of overpriced cocktails, later-than-necessary nights, boring coffee shops, and worse, ordinary conversation. Your dates can smell bad, look bad, and if luck is genuinely horrific on a particular evening, another member of society will court them directly in your face. You never meet anyone at bars, they say. But you might not admit you meet people online either, making your success rate at "Starbucks" awe-inspiring. I met a lot of people at "coffee-shops" when I was using these apps. The potential harassment from sarcastic buddies was too threatening.
Throughout the dating process, it is natural for people to endlessly what-if. What if my date was better-looking, or more successful, or had bigger muscles, or more beautiful hair? These questions lead most people nowhere, but dating sites nonetheless capitalize on curiosity by offering users a vast world of choice, with potential suitors of all variations. Merely keeping up with the electronic chatter can be as exhausting as working through a panic-attack on a first date.
Raya's approach differs from other platforms and their advantages and disadvantages. First, it's relatively unknown and lacks the societal baggage familiar to other sites. The user-base is dissected and curated rather than accepted after a series of standard questions. The typical pitch-phrase "I'm a fun-loving person who loves to laugh" will result in immediate dismissal from the application process. No one hates fun. No one hates laughing. Ultimately, Raya's exclusive acceptance committee drives the app's word-of-mouth buzz. No one knows what disqualifies you from acceptance, and accepted individuals seem alike only by their good looks and mysterious cachet.
A curated club like Raya is a smart idea. Rather than emphasizing an enormous user-base and broadcast-style communication, Raya asks members to keep quiet, discourages screenshots, and is swift to boot anyone who violates the opaque rules. Also, the app costs $7.99 per month. The user-base is probably quite engaged. While folks like you and me may join to date, some will forget to cease payments because they want to know who else is part of the community. John Mayer, professional athletes, and TV news anchors are reportedly there? Sure, keep the subscription just in case. Raya will happily accept your dollars.
Your TV Is Watching You Too
Since moving back to New York in August 2016, I’ve reverted to cable-subscriber status. From 2011 through 2016, I was a cord-cutter. I didn’t care much to watch live sports, HGTV, or listen to the repetitive drone of political punditry. But, I watched plenty of television via terrestrial HDTV signals, illegal downloads, and HBO through the use of other friend's cable subscription credentials. Said another way, my TV was often displaying something.
I view hacking as an inevitability rather than a potential future event. While some sinister individual or organization hasn't stolen my identity, my credit card number has transacted without my confirmation. Sadly, I've probably succumbed to a phishing attack as well. My view as mentioned leads to a somewhat lackadaisical approach to security. My passwords aren’t my birthday, but I also don’t worry about entering sensitive account information on banking websites. At some point, such information will probably flow through the leaky pipes of the internet without my knowledge anyway.
Sensitive personal data passes through my own broadband subscription all the time, but to date, I didn't carefully consider whether my TV was addicted to the internet just like everyone else. My video consumption habits are hardly confidential information, but it now seems evident that what I watch may -- because of my TV -- lead to targeted advertising on Facebook, Instagram, and a variety of other destinations purchasing my watching habits and linking them to my illustrious internet persona. Worse yet, Samba can scuba-dive in my network to look at my other connected devices. The software probably knows I use a ChromeCast and three Wemo MINI smart plugs. Start your hacking!
What I find most interesting about this story is not the transfer of data, but rather than the opt-in process. Or maybe I should call it the opt-out process. The author sources various quotes from Samba employees discussing how clear language is used to ensure consumers understand Samba's software. We know people don’t read the fine print and often click "yes" merely to navigate past screens or interfaces that interrupt their original motivations. I pointed to a similar issue when I wrote about Venmo and Facebook. Users don't pay close attention. The fine-print legalese silently controlling our experiences is an afterthought. The goal of the week: try to identify network traffic coming from my TV.
Computers Will Steal (Some Of) Our Jobs
The linked piece is worth reading more than once. Topics with highly-hyped buzzphrases such as machine learning, artificial intelligence, and the infamous blockchain infiltrate our LinkedIn and Twitter feeds. All of these technologies will steal all of our jobs, leaving us few choices other than proud greeters at Walmart. The internet is overflowing with anecdotes and predictions grounded in little logic and lots of hysteria
Benedict Evans is a smart analyst, and I thought the last third of his essay was the best. That section discusses how machine learning will be used to transform specific tasks from mostly human-oriented or entirely manual in nature to more automated and efficient. These tasks are often narrow in scope such as identifying an object in an image over and over again. Progress around these tasks touched me directly. Part of my day-to-day involves inspecting insurance underwriting data to validate completeness and accuracy. Often, one or both are out of sync, a common problem among complex types of insurance.
Here's an example: you tell me about your home, and I decide whether your home is an attractive risk given the price I will charge for that risk. If so, I offer you said price in return, and you choose to either accept or decline my coverage. In the middle of the transaction, I've asked or sourced whether your home is constructed of wood or brick, has an up-to-date roof, is near the ocean or a fault line, and a variety of other data points. Assuming everything goes smoothly, our relationship can continue in this manner for many years with occasional risk or price adjustments.
Over time, I am keeping track of your home, your neighbor, and perhaps 300-400k others. Key to my business health is ensuring all of the information I have is accurate. This is where machine learning now helps a task that once was very manual and is still human-oriented. Using satellite and low-altitude Cessa flyover data, I can extract the shape and condition of the roof of every home I indemnify using image recognition algorithms. If I find discrepancies between data from the algorithm and data I was provided during underwriting, I can make adjustments to my portfolio and adjust pricing accordingly. Ultimately, image recognition helps me charge my customers more accurately, and avoid unnecessary risks before I've committed my firm's capital for the year.
What will I do now that I don't have to manually thumb through Google street-view images? The universe of potential opportunities is endless. But importantly, as Benedict points out, I can begin formulating other questions for my data or creating new ways to consume and unlock the value from data previously challenging to acquire. In other words, I'll try to come up with a useful solution to another problem, just like any other day.