For those who didn’t know, I’m working with Luna, a blockchain-based dating app. Scott Alexander recently wrote a post about our project, which – along with the comment section – brought up some questions we don’t currently have good published answers for. So here’s some published answers!
This is written assuming you have a basic knowledge of how Luna’s platform works.
- Why a blockchain? This sounds like you are using it just for the hype factor and not because it actually brings anything to the platform. Why don’t you just use in-app points?
Compared to other blockchain projects, we have a greater percentage of our focus dedicated to non-blockchain elements, such as marketing, incentive arrangement, and user interface of the platform. We aren’t just a blockchain company, we’re also a dating app company.
Secondly, I am less interested in the question of the necessity of blockchain and more the question, “Would Luna have a greater chance of success in a world where it used blockchain, compared to a world where it didn’t?” I believe the answer is a resounding yes – based on actual utility of the app, not based on the idea of funding or general blockchain mania.
Blockchain is just a technology, and it’s the technology best suited to the kind of world we want to create with Luna. There’s lots of small reasons why it is suited – the types of transactions done, the flexibility it allows us, the legal situation – but in my opinion the biggest one is psychological – “capturing the value of the network.”
This basically means that the value of Luna’s token directly reflects the amount people value Luna itself. This isn’t the case with fiat – spending a dollar on a mega choco shake at McDonalds doesn’t mean anything special about McDonalds – because you could spend that dollar anywhere.
The same is true with your standard cryptocurrencies, like ETH or LTC. Increased use of a currency also increases the value of all things that use that currency. This is fine for many projects that don’t benefit from a localized network, but for something like a dating app, where there are very explicit network effect pressures, an exchange system that captures value instead of bleeding it starts to make more sense.
Steemit is a prime example. Crudely described, it’s like Reddit but with tokenized upvotes. This style of community shares a lot in common with dating apps – the more users use it, the more valuable it becomes. The most valuable thing on the platform is attention. Interaction with other users is the end goal.
Why does Steemit need a token built on blockchain? Well, it probably doesn’t need one – but taking the token out of Steemit would change it deeply, and for the worse. Steemit has become its own ecosystem, an eddy sectioned off from the general flow of money, and this makes users more invested, increases the sense of community, and rewards early or dedicated users.
- Isn’t this going to be a sugar baby thing?
Integrating money and dating is a really tricky thing to do, and we’ve known from the beginning it’s a delicate balance to strike. We’ve been as careful and as thorough as possible to prevent this becoming anything close to a camgirling, prostitution, or sugar baby platform.
A user can transfer tokens to another through the messaging system only once, ever. The amount they transfer is not determined by them – they pick the most they’d want to pay, but often what they actually pay will be lower. Tokens are necessarily discounted by compatibility rating. The amount a user receives is fuzzed and payout is delayed, so it’s difficult for people to tell who gave them what.
Basically, if someone wanted to use Luna for sugary purposes, I suppose they could in theory, but it would take a lot of effort and be very inconvenient, and they’d be silly for not using any of the other many platforms that do it better.
- There’s already an imbalance in apps, won’t this just favor the rich? Isn’t this just going to be a bunch of wealthy people dominating the app?
Wealthy people will have some advantage in the app, yes – but this advantage is very specific and limited.
On OKCupid (pre-changes), I got an average of around 14 messages a day. The messages were on a spectrum from unique and engaging to a 21% match spamming “hi beautiful wanna ride me” – and the spectrum was definitely skewed towards the latter. If those people want to waste my time, then I’d love to use a system where I at least get paid for it.
And with Luna, we intend for it to be largely those people who pay. Once we have enough data for predictions, we plan on giving discounts proportional to the likelihood of response – so the people sending unique and engaging messages to me will have to commit a fraction of the tokens that all the others will.
So wealthy people will have an advantage on the app – but that advantage will primarily lie in overcoming high cost to message someone incompatible, which means we’ll finally have a system where poor messages are both disincentivized and offset into something palatable.
- I don’t think Luna is going to solve my problems on dating apps.
If you give a bunch of humans the ability to upload photos of themselves and message each other, you will get an imbalance. Any correction of this imbalance means pressure on behavior, which means there is no possible solution to this where someone isn’t doing something they don’t want at least a little bit.
I’m not saying that Luna will completely solve the imbalance or that everyone will get what they want. I am saying that I think it will do a better job at solving the imbalances – while also giving people what – they want than any other platform on the market.
- This system has dangerously high incentive for scammers. How are you planning to prevent fake accounts and bots? What about real people with no intention of actually using the platform to date?
There’s a list of increasingly severe steps we can take depending on the rate of fake users. Here’s a few:
1. Verifying the user account through social media (hooking up with Facebook)
2. Verifying the user account through unique photo means (using live-camera only and taking a photograph under specific instructed conditions, i.e., ‘place a hand on your head while sticking out your tongue’)
3. Requiring additional verification for suspicious activity/accounts via video recording or live video questions
4. Using information from real-life dates recorded with the Luna app
5. Displaying behavioral information publicly on a profile – reply quality, rate, and real-life date frequency (fuzzed for privacy; likely a ratio of dates to time spent on platform).
There’s no foolproof way to prevent someone who is very determined and goes through a lot of work to appear ‘genuine’ on the platform, but we have a lot of ways to make it very inconvenient and reduce the rates to something manageable – for example, identifying users with similar high-suspicion behavior, and grouping them in each other’s feeds and reducing their appearance in others.
- You say Luna’s incentives are aligned, but ultimately isn’t Luna still incentivized to keep users on the platform?
Luna’s incentives are aligned further down the relationship path than other dating apps are – we get rewarded when people respond to messages, and we also plan to collect data from real-life meetings.
The total population of single-and-looking (or poly-and-looking!) is growing; it’s not like we meaningfully reduce our user demographic at all by giving them what they want. I think it’s a better long-term strategy to create a platform that absorbs, matches, and then pushes users off the app. Capturing the churn is what’s important, and for every person with a happy Luna success story we get dozens of friends to hear about it.