- it may not seem like it, but Tinder does try to pick people it thinks are most likely to produce a match, based on-- to be brutally honest-- your attractiveness (as determined by who has swiped for you) and the other person's
- it then orders them approximately from most likely to least likely, but it has to know how often to start over, because eventually as you go down the list you'll eventually start running into people that you're less likely to swipe right on than the people you already passed up
- carefully arranging things so that "super likes" have a reasonable chance of producing a match while also not inundating very attractive people with nothing but super likes from people they will never respond to (which would drive them off the service)
It's a tough challenge and Tinder is not at all a "dumb, simple" dating app.
When the app first came out (or possibly when you first start using it), the first person is/was always a match. Then they got clever about hiding matches deeper in the swipes.
This triggers addictive behaviour because the "match" reward is inconsistent. Though, clearly, before you start swiping, Tinder knows which matches it is going to present to you. It's a delicate balance to judge how long you will spend swiping, when to show you your match and how to make sure you swipe right a few times to keep the funnel loaded.
Perhaps I shouldn't have used the word "dumb" - I do not mean to imply the Tinder strategy or algorithms are stupid. Just meant that instead of prioritizing matching to the best of their ability, they prioritized usability.
At the other extreme, apps like Grindr and to a slightly lesser degree Tinder aren't hurt by doing "too good" a job, since more of their users don't stop looking because they found one good match.
I'd argue that the market of 'people who are dating' isn't going to be dramatically affected by any particular company, and probably not even by online dating in general. They would have to be stupendously effective at creating long-term monogamous relationships to really move the needle there.
Marriage rates have been declining for decades, and serial monogamy seems to be growing (can't find good quantification of that). Successful dating services don't necessarily cannibalize their market.
My intuition is that providing an effective and enjoyable experience is going to provide far greater returns by gaining market share in a relatively inelastic market.
* No Fitbit data == "You're inactive
* No Open source github commits == "you dont like code"
* Few Facebook friends == "you dont have friends"
* Few Instagram posts == "You dont do anything"
I don't think blackballing people due to statistical inferences about their behaviour is any better than harassment.
> I’d guess the findings were racist: OkCupid statistics show that even though people say they don’t care about race when choosing a partner, they usually act as if they do.
It's weird that wanting to produce children with someone of the same ethnic background as oneself is seen as racist. It seems quite unremarkably normal to prefer folks from a similar background. There's nothing wrong, of course, with those who choose someone different than themselves, but I think they're the outliers.
What? Harassment is active disruption of your life. Hiding you from someone's search results or "compatibility" list is not. It's not a great future, but let's have some perspective and not equate apples with oranges.
I think wanting to produce children with someone of the same ethnic background as oneself is very mildly racist, but it's not nothing.
Netflix recommendations are utterly bad.
Musk et al. warn us constantly about "AI" and its dangers, but so far AI doesn't seem to work very well. Translating from Italian to English with Google Translate results in complete gibberish; Amazon (the other recommendation engine everyone mentions in that kind of discussion) is only ever able to suggest products one has already bought.
It's probably hard to judge the efficiency of those dating systems since there is no control, and people want them to work. It would be interesting to compare a "sophisticated" dating engine with a more random one and see if there's any difference in outcome.
During my divorce, I intentionally sought to not allow past relationship data to influence whom I hooked up with.* I had been with four people. Three of them were blonde. Three of them were born within a few weeks of me. All four of them were either in the military at the time or joined it later (because we were both just 17).
I didn't hate the man I was divorcing. I didn't think I had missed by much on finding a great match for me. I was trying to figure out how to sort the wheat from the chaff.
For me, a dating app that said "You clearly like blondes. Here are all the blondes." would have been the exact opposite of what I wanted. I wanted to make personal connections that were not biased by shallow details of that sort. I succeeded in that goal and it was a growth experience.
So, to my mind, a dating app that would be this thorough is the dating equivalent of redlining colored communities and not approving mortgages there. I say that because one of the issues was that I had been molested as a child. So, my feeling is that if you limit future relationships based on passed behavior, you are saying that you are 100% fine with painting abuse survivors into a corner they can never, ever escape. And that doesn't sit well with me.
I generally like to give life room to surprise me. I have gone out of my way to let men surprise me with being better people than my bad past experiences led me to believe I could find in men. The last thing I want is someone or something essentially telling me "So, we hear you like abusive men who will shit on you. Here are your matches."
Yeah, no. I never asked for that. But finding my way out of it has been a long, strange journey because, seriously, society doesn't yet have a good play book for how on earth you do that. So, you very much have to roll your own.
* http://micheleincalifornia.blogspot.com/2015/10/reducing-bia...
It also ignores how attraction works. Sure, someone may find blond people attractive, but it's seldom a deal-breaker. These algorithms tend to select on characteristics that are easy to assess, discarding many potential partners who'd be more suitable in other ways.
It's starting to approach that point in the adoption curve where skepticism fades and people think of it as the solution for (or threat to) everything, rather than as a tool that is good at some things and not at others and that has characteristics like accuracy. With AI and the widespread blind acceptance of our surveillance society, that could be a dangerous point of view for the public to embrace.
Intersting thing with this is I have 4 different twitter accounts for all of my different personalities:
- political
- work related
- life(dog) related
- shit posting
I kind of want to see the different matches I’d get based on my tweet taxonomy.
They're measuring compatibility on language analysis of your tweets? This seems inefficient and unnecessary.
Wouldn't it be both simpler and more effective to analyze who you follow for compatibility? In regards to feeds themselves, retweeting the same article could be a great talking point and measure of similar interests.
Then again, maybe I'm overestimating the intelligence (or integrity) of dating app purveyors. After all, 'real' people are so hard to monetise...
For example, I was an avid user of IRC back in the late '90s. That medium formed my social experience and those people became many of my life-long friends. These days many of my friends from that time (including myself) don't use social networks much at all; Facebook, Instagram, and Twitter included. We've been there, done that, and are finding human engagement without social networks to be more satisfying. Consequently, my contemporary relationships are not significantly maintained through social networks and whether someone is or is not on Facebook or Instagram has no bearing on whether my relationships are maintained.
A wise friend of mine once said "Facebook is for people you don't know!"
Not wanting my insurer to have that data presumably to them means I'm unfit, rather than skeptical of what they're going to do with it
I'm joking of course, I do not condone insurance fraud.
At one end of the spectrum, insurance companies, or dating apps, have essentially no data. So every person is charged equally, which I think is unfair (some might disagree).
To improve, companies collect data via a questionnaire. The obvious problem is people lie -- often unintentionally.
The next step is to use genuine data. And, there are certainly issues with that.
It's just not clear to me that the issues outweigh the problems with the alternative scenarios.
As someone who worked for an online marketing company and designed its tracking system, i can tell you that’s the most frightening part of it. Everything is logic,and is a logical evolution or improvement of the previous state.
Every intrusion of foreign company into your private life will be done for your own good, or at least for the benefit of the society. You don’t even need an evil genius at the head of some powerful tech company for 1984 to happen, because it’s simply the logic of the whole system we’re living it, that’s meant to go this road.
Or the way the world is going the next step is a startup that sells an arm that tricks smart watches into thinking they are on a human arm that is exercising regularly...
But from the perspective of people heavily invested in those platforms, you don't. And as the number of people on these monoplatforms trends towards 100%...
Well, I dunno, maybe it's just big cities that are like that.
i always like to bring this up when people mention algos only being able to shuffle the past - http://www.bbc.co.uk/blogs/adamcurtis/entries/78691781-c9b7-...
i would think music or photos or writing would be the first place where algos could take WHAT you like about something and find matches. Just liking or upvoting isnt enough in my experience. But with music if I can say I like the lyrics, or beat, or instruments or vocal tones..
> One Target employee I spoke to provided a hypothetical example. Take a fictional Target shopper named Jenny Ward, who is 23, lives in Atlanta and in March bought cocoa-butter lotion, a purse large enough to double as a diaper bag, zinc and magnesium supplements and a bright blue rug. There’s, say, an 87 percent chance that she’s pregnant and that her delivery date is sometime in late August. What’s more, because of the data attached to her Guest ID number, Target knows how to trigger Jenny’s habits. They know that if she receives a coupon via e-mail, it will most likely cue her to buy online. They know that if she receives an ad in the mail on Friday, she frequently uses it on a weekend trip to the store. And they know that if they reward her with a printed receipt that entitles her to a free cup of Starbucks coffee, she’ll use it when she comes back again.
<...>
> Pole applied his program to every regular female shopper in Target’s national database and soon had a list of tens of thousands of women who were most likely pregnant. If they could entice those women or their husbands to visit Target and buy baby-related products, the company’s cue-routine-reward calculators could kick in and start pushing them to buy groceries, bathing suits, toys and clothing, as well. When Pole shared his list with the marketers, he said, they were ecstatic. Soon, Pole was getting invited to meetings above his paygrade. Eventually his paygrade went up.
<...>
> About a year after Pole created his pregnancy-prediction model, a man walked into a Target outside Minneapolis and demanded to see the manager. He was clutching coupons that had been sent to his daughter, and he was angry, according to an employee who participated in the conversation.
> “My daughter got this in the mail!” he said. “She’s still in high school, and you’re sending her coupons for baby clothes and cribs? Are you trying to encourage her to get pregnant?”
> The manager didn’t have any idea what the man was talking about. He looked at the mailer. Sure enough, it was addressed to the man’s daughter and contained advertisements for maternity clothing, nursery furniture and pictures of smiling infants. The manager apologized and then called a few days later to apologize again.
> On the phone, though, the father was somewhat abashed. “I had a talk with my daughter,” he said. “It turns out there’s been some activities in my house I haven’t been completely aware of. She’s due in August. I owe you an apology.”
Now my memory isn't great, but I think there was a fair amount of uproar of this and Target scrapped the program. There was a line crossed between "useful" and "creepy" somewhere along the way, and it would be my guess that Amazon seeks to avoid that. I'd fully expect them to have the data and engineering capability to create a very powerful recommendation engine (the Target article is five years old now...) - but the expected value isn't there.
But what about others? What would be creepy about excellent Netflix recommendations for example?
GP is clearly not suggesting that a dating service can be so effective that it depletes the market of single people. They're suggesting that a successful dating service won't retain customers, insofar as customers who find their life partner no longer need dates.
My point is that it's important to consider all of the trade offs.
The Algos have pros and cons.
When I hear people excited about slack, I see it as the modern day irc client, also focused around communication.
ICQ/AIM took a big chunk out of irc, but it still didn't allow people to connect in channels.
Now we have networks where people aren't communicating as much as sharing and interacting with moments.
The uncompromising and unquestioned loyalty that exists from having a buddy in another city is something social media (and maybe even dating sites) could learn a thing or ten from.
I'd say online message boards (phpbb, etc) in from 99 to 2007 or so were similar networks. There is something about semi long form communication that can't be liked, tweeted, shared, etc to be improved.