A blockchain to stop sybil attacks

Fake accounts are ruining the internet. Social media in all its forms is a carefully orchestrated dance of bots interacting with real people. These fake accounts and bot networks have the sole purpose of extracting value without creating anything useful.

Some blockchains are attempting to solve this problem through understanding the network. For example, if I know Sarah and Sarah knows Bob, I can trust Bob because I know him through Sarah. Similarly, if Bob creates a bunch of fake accounts that only know Bob, it’s easy to prune the network of Bob and all of the fake accounts that only know Bob.

Such attempts although logical, do not stem the economic incentives to create bot accounts. If you build a fortress and hide gold in it, someone will figure out a way to break in and steal it. Blockchain works as a store of value because it fundamentally changes the economic incentive. Everyone knows 51% attacks (as well as others) will always be possible in bitcoin and other POW systems but they get increasingly expensive as the network grows. As bitcoin grows, it becomes unfeasibly expensive to attack the network in meaningful ways and doing so risks the value extracted.

As blockchain continues to prove its value as a currency and a store, the next wave of DAPs seems to sidestep the Sybil problem. You’ve probably already noticed on Steemit how many accounts repost content without thought and try to manipulate the system without attempting to create value.

As long as there is an economic incentive to create fake accounts & bots, bad actors will continue to create fake accounts & bots.

Problem proposal & System Requirements:
From an economic perspective, we must design a system such that creating 1 account creates value for a user and creating > 1 account imposes an economic cost for the singular owner.

More formally:

1. Having 1 account produces value, e, to the user
2. Having n accounts sums to a value < e that strictly decreases as n grows.

Existing solutions:

1. Make user accounts cost money aka micro-transactions & sign-up fees. This system meets 2 but does not necessarily satisfy 1. The problem is that having an account must provide value. If the cost of creating or using an account is > e, the network will never grow and users will leave. Friction is the enemy.

Proposed Solution:
Let’s call the network, Singleton and the protocol, Proof of Existence.

Proof of Existence:

1. The network work would store two actions on the blockchain. Create and destroy.
2. All users as rewarded for every block that appears between create and destroy blocks.
3. Rewards scale inversely exponentially to the number of created users.
4. Validating creating / destroying out of users is rewarded.

This could be built on top of Ethereum and used as the basis for user management to DAPs.

A couple observations:

1. Since the reward for existence is exponentially decreasing, a human would never want to have multiple accounts as long as the exponential factor, d is less than 1/2. For example, if by existing as one account I get a block reward of e with a d of 1/3, creating another account would give me a reward of 2 * ( 1 / 3 * e).
2. If someone tried to attack the network by DDOSing creation of accounts, the reward for existence would get very small quickly.

I'm hoping to start a discussion on this top of solving Sybil attacks because many of the articles I've read so far only talk about how they will fix/patch/make hard problems instead of fixing the economic incentives for them.