Why luna?

By Vinay Gupta, formerly ethereum project

I’m a person who is usually known for chasing big, important problems:

climate refugees, critical infrastructure, pandemic flu, environment and so on. Why, then, have I been lugging around an idea about how to fix online dating for a few years, talking about it, and trying to get people to see different ways of thinking about it?

The answer is because the best things in life are free. There’s no significant economic or social cost to people being with the people who will, at any given time in their life, make them happiest.

And this is essentially a search (and possibly scheduling) problem. Let’s deal with the search part first. Humans in traditional societies knew about inbreeding. They evolved elaborate ceremonies with and traditions for making sure that their kids married outside of the tribe, even if there were serious difficulties (like language barriers) in doing so. More than one agricultural society held no-holds-barred festivals in which you could arrive with who you liked, and leave with somebody else, and nobody was really allowed to give you a hard time about it.

This stuff didn’t happen for no reason. It happened because getting the biological imperatives right mattered. Presumably the cultures which didn’t have these kinds of mechanism for keeping the gene pool stirred died out, or perhaps human instincts always settle on some kind of solution. 

But, one way or another, complex social arrangements around ensuring exogamy have always been the norm. Even in arranged marriage cultures like my own (about half of my cousins had arranged marriages) people would go hundreds of miles for a match in many cases. A lot of time and effort went into these things. Dating wasn’t just a case of whipping out an app, or heading down to the local bar district and hoping that common sense wasn’t picking up the phone tonight. No, it was a case of long journeys and adaptation to a foreign culture and diplomatic relations with the neighboring tribes, and all the rest of it. These things involved serious effort.

But now the general model is to make everything as easy as possible: birth control and barrier contraception, no social pressure to make relationships lasting until quite late in life (or ever), and in some lucky countries social care systems so good that single parents don’t have lives any harder than married parents all contribute to a general sense that sex and relationships are driven by convenience rather than being in some sense the centre of life.

We have basically de-labored the process, and the results don’t seem to be making people happy. I’m not comparing that level of happiness with some imaginary “good old days” — just noting that, on average, the entire thing is filled with angst. 

So why do I think the blockchain can help? 

The answer is twofold. Firstly, there is the agency problem. If you are on a dating site that gets paid when you see advertising as you search profiles, you are going to see a lot of profiles. They might withhold useful search features or pair you with a lot of people you won’t like, but who aren’t quite so obnoxious that you put the site down, and just feed you a smattering of great people to keep you interested. 

Or they might have seas of appealing bots whose job is to keep you clicking, while delightful real people are just out of reach, tucked away in a corner of the database their algorithms just won’t quite allow you to see. And then there’s the actual performance of these systems — you’ll find a lot more analysis about this in the whitepaper, but the results are quite discouraging. The bottom line is that these sites don’t work directly for the best interests of their users — there’s always a split motive, and this particular aspect of our lives is too important to let the agency problem interfere with our happiness. If there is to be machine help for people to find partners, the machines have to be working strictly on behalf of the people they are connecting, not hiding their algorithmic intentions behind complex veils made to sell advertising. 

The second problem is more subtle. Dating sites bake in all kinds of hidden assumptions. Take age: does it really matter, past the age of legal consent? Hard to know. For a very long time, asking people their age was considered impolite. Now, it’s right up there, front and centre. Does that make for more or less successful relationships? How could we know? How could we watch? How could we measure? Same thing for photographs. Of course looks matter, but are we selecting for nice looking people, or good photographers? Maybe dating people who are really good at selfies does genuinely result in better relationships — or maybe not!

The bottom line is that we can’t really tell, because all of this useful data is locked away inside of the matching engines of the data sites. OkCupid’s blog series on what they’d learned from their data is absolutely fascinating, but it leaves me wanting more: much more!

How could we design matching engines which genuinely produced human happiness?

The answer would require data — both big, and open, and we would need to know something about the long term outcomes of various matches made, sometimes decades before. An algorithm which produces 5 brilliant years, then a string of massive ugly divorces might be suboptimal for reasons we can’t possibly guess right now — and if that seems like an improbable outcome, let’s remember that computers can recognize images better than people now. Maybe they can recognize compatibility too.

But for us to ever find out, we are going to require an open data ecosystem around computer dating. Blockchain is an integral part of that — it’s what pays the bills to do the science and, in the case of Luna, it nicely and accurately solves one of the key problems in the computer dating arena: cut-and-paste messages spammed over huge numbers of people, resulting in an ever-lower number of good quality genuinely interested messages, hidden in an ever larger sea of dating spam. 

Just getting rid of that dynamic once and for all would be a great result, but I think that Luna offers far, far more.

By establishing the decentralized paradigm in dating, Luna helps to remake dating culture.

Luna is not a service or a place, like Tinder or a bar. Luna is a method, and a method which can be continually improved using techniques like A/B testing, until it is genuinely producing better lives for people.

Because blockchain techniques allow for sophisticated tools to be developed to align economic interests between (say) search algorithm designers and individual users, or between users and other users who don’t like spam (i.e. everybody), the possibility exists to not only solve the questionable agency of the current generation of dating app providers, but to create positive agency to do something really, really new. We could pay the best people in the world to design algorithms to match other people, and make them happy.

Luna doesn’t start with algorithm markets. Designing the necessary markets and mechanisms to genuinely reward people for doing the necessary science to get more optimal online matching is an uphill job, and the Luna team know that. So they’re starting small: a token economy for dating. What does this token do? It streamlines attention: messages sit in an inbox, and the messages that people paid most to send sit at the top of the inbox, with some modifiers which will be A/B tested along the way. That doesn’t seem like a big deal, but it is. Once that flow of value is established — that these messages matter, that they are indeed precious cargo — economic incentive mechanisms can be designed to help people better target their search.

Once the option of just spamming messages over half the site is gone, once you can’t right swipe on everybody, the market exists for the design of algorithms which help you spend your time and energy talking to the people who are most likely to find you the person they wanted to meet. We can’t climb that hill towards genuinely better matching without an underlying tokenized economy to pay for the research, and we can’t get better results without that research.

So, phase one, tokenize all the things, and use the simple brute force method of charging fees for communication to prune spam. To keep the system fair, the fees go to the person you’re communicating with (minus a fee). Then, later, offer ways to take commission on introducing people using a variety of matching algorithms, ideally in a competitive array, and use that mechanism not just to prune the bad messages, but to actively steer and guide people towards each other based on criteria they might never consciously be able to articulate, but that sophisticated machine learning algorithms may able to sense.

This is a big, bold, long term vision, but it’s a good one.

If computers are good at anything, it is search. We aim to establish a market which sets up the right conditions to train them to do something important and something new: to train them to search for love.

October 17, 2017
© 2018 Gravity Analytics Ltd. All rights reserved.