We already have a pretty good Sybil resistance against duplicate accounts with the face verification. But one thing we don’t really have good resistance against: bots that automate the claim every day.
An idea discussed early on in the project was that the daily claim is a form of “proof-of-need”. We want to allow anyone to claim, but with the assumption that the daily claim requirement would discourage people without a need to come back daily. It’s my assumption that anyone who can afford a bot and can spare the time enough to set one up probably don’t have the “need” for UBI. Such bots also do a disproportionate damage to the system, by being way more effective at draining the reserve than any normal human.
My proposal is simple: Add a captcha requirement before claiming. It doesn’t have to be a daily requirements, it could be a requirement that is added every 7 days for example. This way a daily claimer would only have to do the captcha one per week, and an infrequent claimer not often at all.
The problem would be that a captcha requirement can only be enforced in-app. It’s very hard to enforce (I assume) for CLI-based requests (which is what the bots use). But what we could do with those is simply filter them out and apply post-claim rules. For example, if a user makes more than 6 consecutive CLI requests then they get de-whitelisted. Or something else less drastic, perhaps they get put on a gray list which would make them unable to claim again until they’ve done an in-app claim which would whitelist them again.
If we were to hypothetically assume that the bots make up 20% of claims, then if we stopped them it would mean that a daily claim could go from 60 to 75 value for everyone else. (Disclaimer: I have no idea about how many bots there are. But I’m sure you clever guys could analyze and estimate the number.)
Anyways. Just a thought. I’m bored on a long flight home after seeing my fave GoodDollar peeps in person, and so naturally my head start spinning in that direction again.