(On Legitimacy of Blockchains - 18)
The whole point of leveraging the idea of P2E (Play to Earn) as the primary means of maintaining a state of consensus is to ensure that the blockchain's protocol is so unfathomably complex, that it is virtually impossible for even a highly sophisticated general-purpose AI system to exploit in order to get ahold of unfair advantage.
For ease of conveying the subtle beauty of the technology behind this, please let me explain the concept based on a slightly metaphorical analogy. Suppose that there is a blockchain which is being operated by a unique type of PoW (Proof of Work) consensus mechanism in a sense that it assures the presence of a node's computational work not by the fact that it has found a special input value to a cryptographic hash function (e.g. SHA256) which yields an output value with a special binary pattern, but by the fact that it has discovered a sequence of player inputs inside the game of chess which yields a specific result. Whenever a transaction gets issued, for instance, it can create a round of chess in which one of the virtual players generates a sequence of moves that are pseudo-randomly determined by the transaction's hash code. As soon as the game gets broadcasted throughout the network, miners can competitively try to come up with a sequence of moves of the opponent player which results in the opponent's victory. Anyone who first succeeds in doing so will receive the mining reward.
Now, of course, such a mining rule is easily exploitable by any software that is capable of traversing the game's tree of possibilities and quickly discovering one of its branches that leads to the win-condition. This is because the game of chess is merely a set of logical rules which do not change at all, which means the shape of the game's possibility space will never undergo any modification. It must be fairly easy for a computer, then, to find an optimal path upon the surface of such a static landscape.
Let us now imagine that there is yet another blockchain in which the mining logic operates not on a simple game with a collection of fixed rules (like chess), but on a multiplayer sandbox game in which the rules themselves can be periodically amended as a result of virtual votes being cast by its players. A PoW algorithm which runs upon such a dynamic system must be orders of magnitude more resilient to hackers than a static one because its possibility space is a fluid-like surface whose control points fluctuate at random moments in time due to random actions triggered by its human participants. Even the most advanced machine-learning program cannot decode the nature of such a living galaxy of chaos, governed by countless human souls each of which is a indecipherable body of complexity in its own right.