Dating apps/sites that employ the strategy of "get as much data as possible and match intelligently" have failed to compete for users against Tinder, the dumbest, simplest dating app that hardly does any intelligent matching at all. This implies that it's not the sophistication of a dating app that matters, it's the simplicity and ability to attract a userbase that in turn attracts a larger userbase etc in a virtuous cycle.
To me the frightening thing is that people seem to be unable to differentiate between no data and the negative case. For example<p>* No Fitbit data == "You're inactive<p>* No Open source github commits == "you dont like code"<p>* Few Facebook friends == "you dont have friends"<p>* Few Instagram posts == "You dont do anything"
When people talk about AI threatening humanity I suspect folks imagine The T-2000 model Terminator. But here is the real thing; AI is going to change our society in a profound and slow-moving way, very different from our perspective of how war and soldiers change power structures.
> But in the future, apps could identify sexists/racists/homophobes by their social media activity and preemptively blacklist them from joining. (Maybe this would aid the industry’s problem with harassment, too.)<p>I don't think blackballing people due to statistical inferences about their behaviour is any better than harassment.<p>> 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.<p>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 <i>normal</i> to prefer folks from a similar background. There's nothing wrong, of course, with those who choose someone different than themselves, but I think <i>they're</i> the outliers.
> <i>Already, some apps do this by learning patterns in who we left and right swipe on, the same way Netflix makes recommendations from the movies we’ve liked in the past.</i><p>Netflix recommendations are utterly bad.<p>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.<p>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.
One of the problems with this idea is that it kind of paints you into a corner. It doesn't allow you to aspire to be the person you desire to be. It predicts that who you are is static and then it helps make that a reality.<p>During my divorce, I intentionally sought to <i>not</i> 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).<p>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.<p>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.<p>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.<p>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."<p>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.<p>* <a href="http://micheleincalifornia.blogspot.com/2015/10/reducing-bias-in-dating-small-personal.html" rel="nofollow">http://micheleincalifornia.blogspot.com/2015/10/reducing-bia...</a>
What I find most disturbing is the viewpoint underlying this article, an emerging faith in AI and a magical ability to divine things about the world, with no discussion of its strengths and limitations, its false positives, false negatives, true positives, etc.<p>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.
Married, so I probably won’t be using that site anytime soon. I did meet my wife on OkCupid and I was doing the Ng ML class at the time and studying NLP and wrote a bit of my own “crawlers” to weed out people, find matches, and draft letters.<p>Intersting thing with this is I have 4 different twitter accounts for all of my different personalities:<p>- political<p>- work related<p>- life(dog) related<p>- shit posting<p>I kind of want to see the different matches I’d get based on my tweet taxonomy.
> it’s paired with the language processing company Receptiviti.ai to compute the compatibility between me and its user base using the contents of our Twitter feeds.<p>They're measuring compatibility on language analysis of your tweets? This seems inefficient and unnecessary.<p>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.
This is so wrong. Your partner is someone that balances you, not a mirror of you. Sure, you need a lot of similar interest, but also differences to get a relationship work, and those I don't yet see in these algoritms, not even remotely.
So I signed up to evaluate the application but after giving the platform access to my timeline I do not see any of the breakdown the author is talking about. Is it because I'm using the web interface?