Proof of Stake

Every blockchain needs a way to agree on what happened. Proof of Work (Bitcoin’s model) solves this by making validators do expensive puzzles. Proof of Stake solves it differently: validators post a large amount of the network’s own token as collateral, and the protocol randomly selects them to propose and approve new blocks in proportion to how much they staked. Misbehaviour costs them their stake, so economic rationality pushes them toward honesty.

The security model is economic, not computational. A PoW attacker needs to control more than half the hardware hash rate to rewrite history. A PoS attacker needs to control more than a third (for most modern PoS chains) of the staked supply to halt finality, and more than half to rewrite it. The economic cost of attacking a PoS chain is the market value of the tokens they would need to buy plus the slashing risk if the attack fails. For a chain with a liquid multi-billion-dollar market cap, that’s usually prohibitive.

PoS has practical advantages over PoW that matter for DeAI. Energy consumption is two orders of magnitude lower (Ethereum dropped its energy usage by ~99.95% when it switched from PoW to PoS in 2022). Validators run on commodity hardware instead of specialised ASICs, which means the barrier to running a node is accessibility rather than a capital arms race. And slashing gives the network a way to punish specific bad actors that PoW chains can’t, because there’s no stake to take.

The weaknesses are real too. Stake concentration is the big one: wealthy token holders can accumulate disproportionate control over block production and governance. Early PoS chains often have long-tailed distributions where early participants hold enough stake to dominate validator sets. This is why the Nakamoto coefficient (how many validators it takes to control a third of stake) is a more honest measure of decentralisation than the total validator count. The OYM Freedom Score’s Infrastructure dimension captures this distinction directly.