Fuseki.net

Crypto TFR multi-level marketing scheme

A Bitcoin millionaire has two kids. He doesn't want to spoil them, and he doesn't want his name to die out, so he implements a smart contract: Starting even before he dies, 1% of his net worth will be given to his kids annually, split between them proportionally to how many children they have.

Generation 1: originator - commits X money
Generation 2: his two kids
Generation 3: his grandchildren.

As soon as a grandchild is born, their gen2 parent gets 100% of the annual 1% disbursement. Inheritance starts before the death of gen1.

Why would someone do this?

They may have a crazy set of beliefs. The key isn't so much why, as evaluating whether this is a stable or growing strategy. Why do deer eat and have fawns? It doesn't really matter; most of their emotions about it are after-effects; the key point is that it survives. Strategies don't have rationalizations when they're invented; the first multicellular life didn't start with a thesis "it's good to cooperate". The only reason we observe it is because it evolved into us.

Requirements:

crypto oracle for validation of identity - the hard part.
gen1 precommits to doing this by sending their entire fortune to the smart contract.
smart contract design.
Also, this would be designed to be infectious: when all of gen2 dies, the gen1 money has a final distribution based on the proportional reward at that time
this isn't coming out as cash - it goes into new smart contracts, taking each member of gen2 as the founding member of a new contract-dynasty.
alternatively, this could actually be run multi-level - gen2 is rewarded while living for having their own grandchildren.

Since it's unlikely to ever exhaust the money, it would either go on forever motivation to

Failure states - oh, so many, but they make for good plots

An overpopulated world where most computation is locked up in mining, and most people or a significant class makes their living by being rewarded for having kids. This would be somewhat alleviated by it being slightly zero-sum - perhaps if more kids increases the disbursement rate rather than the proportion, it would put a cap on the rewards from this kind of strategy
perversions designed to trick the oracle's methods of determining ancestry; after a few generations ages of generations don't necessarily match up and it could be hard to determine what's going on. i.e. if it were multi-level, there would be advantageous marriages which consider the number of generations from the founder. Eventually there would be people from gen4 and gen8 alive at the same time (due to differing parental age), and they would either be preferred or discriminated against, depending on the relatedness calculation made by the smart contract.
other even worse perversions centered around the mismatch between the conventional meaning of a child - someone loved, cared for, and educated, and the naive way the oracle probably determines ancestry.
like many other crypto ideas, the problem is in forcing a correspondence between reality and what's in the blockchain.
The true failure state is that we only reward temporary survival at one generation. Ideally we'd want to reward it multi-level - reward people for having & raising kids who don't just survive, but go on to thrive and have their own kids. Or at a more human level, kids who do great things. So you could make the reward systems proportional to other achievements - such as the hash rate proven by the child. This has risks along the optimization/exploitation axis. The more you fine-tune your measurements, the harder to do correctly and more exploitable they are. The "Friendly AI problem for TFR-increasing crypto contracts"
another failure state is if the oracle attempted to use genetic similarity to the founder to decide relatedness. This would lead to extreme cloning - which the founder might consider beneficial. But also even worse, if the oracle is dependent on a specific genetic test, being able to trick that test is now extremely valuable.

Why is this even a thing?

I think the difficulties here are because identity isn't real. It's extremely hard to define and track in a rigorous way. Things like hashrate or existence at all are quite a bit easier to verify.