Trying to find a way for lightning to spread like wildfire and focused primarily on Africa (but this could work in other southern continents).
This can and will happen. It's amazing when listening to Twitter spaces created by groups in Africa how quickly things are progressing and what a great match lightning is for our problems. Somehow the vibe feels different. There's a lot of positive energy, no fed talk, no price talk, just people trying to figure out a way to help their fellow citizens. The focus is more on how to spend sats, rather than just stack sats, which reminds us of a certain whitepaper from a while back.
When we started this experiment (Lightning Village), it was a bit tricky to figure out which wallet to use. In order to understand this challenge, you first need to imagine the type of users we would be dealing with. Unreliable internet, electricity is a luxury, etc, etc. All the different trade-offs (more about them later) had us spinning round in circles.
So we looked at what builders had done so far in this region and thought by far the best product out there was of course Machankura which everyone here should have heard of. However, it had two main issues. (1) It's custodial and (2) for lightning to really take off massively, we felt the UI needs to be on a smartphone, even if it's an old model, with very low spec.
During this exploration phase, we've experienced lots of issues with wallets. If you do a "mom test" and ask your mom to send you some sats, you'll know what we mean. It's not just "how to receive/send", it's everything from where to get the sats in the first place. Since you can't send natively on-chain to a lightning wallet and not even the most maxi of all exchanges Swan, supports lightning. There's too much digging that needs to be done just to get started.
If you're a bitcoiner, you'll figure it out, of course you will. When there's a will there's a way, even for non-technical users. But for those of us trying to facilitate mass adoption, the goal is to find an end-to-end flow that we can introduce to a small commuity that regularly trade with each other, then incubate it for a while (iterating, learning and adapting the education). Once everything is flowing, the circle of users can get bigger and bigger.
If at each stage, the person trying to introduce the next one, gets asked questions like, I tried to send my relative some bus fare and they can't create an invoice because their phone is not charged, they'll just give up and choose the "pain" of using fiat cause the money is needed there and then. So it ends up becoming a thing for different large groups of bitcoiners spread over Africa (mainly in large cities) to use and solves a few issues here and there, but never really takes off on its own.
So, what has all this got to do with custodial wallets? Well, it's important to set the context first. When thinking about the problem that needs solving, which is getting villages across Africa (who actually really need lightning) to use it, what matters the most is the trade-offs and how they impact the UX.
Before going through the trade-offs, this is what we learned about them in this context:
- the trade-offs in an African environment are slightly different from the trade-offs in the western world
- the trade-offs for a Bitcoin wallet are also different from the trade-offs for a Lightning wallet (anywhere in the world)
Why? Well, first of all, what are the main trade-offs?
a) self-custody
b) privacy
c) no-kyc
d) LNURL support
and e) requires running your own node or knowledge of how to connect to one
Let's focus on (a) self-custody, since that's the title of this post. Starting with 2a, for bitcoin. Imagine a custodial bitcoin wallet. So you buy bitcoin on an exchange (excluding Relai). You don't even know if it's there because you don't have the address. All you see is a number on a screen. That's rug-pull and counter-party risk 101. But a custodial lightning wallet (1a), well first of all, the "node provider" (not exchange) is not "selling tokens". It's not the same type of risk. Secondly, you're not storing large amounts like you would on a bitcoin wallet. So if you get rugged and lose your lunch money, mom sends you some more tomorrow, big deal. Of course, 20 years from now when a single SAT is worth a hell of a lot more, it will be a different story, but we're talking about present time. Also, the most downloaded wallets are well publicised in the community. Scammers don't really get much traction. If people's funds were to disappear from a wallet one day, within hours everyone will know about it and that will be the last time anyone used that one. Remember these users will be spending their sats day to day. Phase 2 next year will be the savings account (btc self custodial).
More importantly, we're educating and onboarding users that have never even heard of bitcoin before. It's all about, 'here is the wallet, now go'!! Watch @JoeNakamoto's videos on Twitter and you'll know what we mean. The point is, once a small community is onboarded and a circular economy has been created, there will be plenty of time to move up the "trade-off ladder".
So we started off with Muun which we thought was great and wrote a nice blog post about it. After using it for a while, we noticed it sucked for our needs. We even helped them fix a bug, hats off to them for their quick response. This is what one of their devs sent us (after the bug fix):
Muun: "You are trying to scan an LNURL-pay which Muun doesn't support at the moment. Muun mistakenly reads the QR code as an LNURL-withdraw, which we do support. We'll fix it. Thanks for reporting it."
That's where we started understanding why lighting still has a bit of way to go. The difference between LNURL-pay and LNURL-withdraw, that's way too technical for your average, everyday users.
We then realized that Wallet of Satoshi actually works pretty well. The only reason we had dismissed it initially was because it was custodial. So we've changed our mind now. Or we've realised that this is a small trade-off which is well worth it to facilitate mass adoption.
There is one UX problem which is that freaking long LNRUL. If only someone could copy their wallet concept where the onboarding is so fast and simple, but register a 3-letter domain, we would switch immediately. Luckily, there's no "wallet loyalty" required for the users and no switching costs. They can just transfer their balance to the new address or spend to 0, then download a new one. The best product will win in the end and other wallets will continue to be great for other types of users with more sophisticated needs, win-win for everyone.
The most important trade-off for us is actually 1c) no-kyc. In the west, the majority of people don't think much about kyc, because its part of life and they're used to it. In Africa, kyc bureaucracy excludes the majority of the population from financial services. This one's 100% non-negotiable for us and also what makes btc and lightning so different. In the btc world, custodial wallets almost always require kyc. Lightning has finally made btc usable and the foss movement will truly change the world.
Then (1d) LNURL is important only because it's the way to send sats without generating an invoice. The invoice-generating thing is a shop or store use case. In the majority of cases, you just want to send someone money the way you've always been able to do it, first with a lot of pain (WU), then with much less pain (WR and MoMo), but never before did you need the recipient to first generate an invoice. So (d) is more of a UX issue, but UX is super important for our mission.
So for now, we'll go with Wallet of Satoshi and if you have any ideas, feel free to reach out, our DMs are open.
Next challenge - what's the easiest, non-KYC way to go from fiat to sats (not fiat to btc)?
keysendandAMPatomic multi-path payments (which obviously fan out when needed if channels aren't wide enough).