What are miners and validators actually doing, and what are they paid?

Both mechanisms answer the same question — who gets to add the next block, and why should we believe them? — and they answer it with two different resources: energy or capital.

Proof of work: guessing, at scale

A Bitcoin miner does something unglamorous and precise.

1. Collect pending transactions from the queue (the mempool), preferring higher fees.
2. Build a block header containing the previous block's hash, a summary of the transactions, a timestamp, and a nonce.
3. Hash that header with SHA-256, twice.
4. Check whether the result is below the target number. It almost certainly is not.
5. Change the nonce and go back to step 3. Repeat, billions of times per second.

There is no shortcut, no clever mathematics, no way to reason toward the answer. The only method is guessing, and your chance of finding a block is simply your share of total network hashing.

Difficulty adjusts every 2016 blocks — about two weeks — moving the target so that average block time returns toward ten minutes regardless of how much hardware has joined or left. This is why the ten-minute cadence has held for over fifteen years while the hardware behind it changed beyond recognition.

What the winner is paid: the block subsidy (currently 3.125 BTC) plus every transaction fee in the block. Nothing else.

What secures it: rewriting history means out-hashing the entire honest network continuously. The chain is expensive to attack precisely because it is expensive to run — the electricity is not a side effect, it is the mechanism. That is both the criticism and the design, and you cannot remove one without the other.

Proof of stake: capital at risk instead

Ethereum switched from proof of work to proof of stake at the Merge, 15 September 2022. The resource changed from electricity to locked capital.

• A validator activates by staking 32 ETH.
• The protocol pseudorandomly selects a proposer for each slot, and committees of validators to attest to what they see.
• Honest, timely attestation earns rewards; being offline loses a small amount steadily.
• Slashing — forcibly removing a portion of the stake and ejecting the validator — is reserved for provably contradictory behaviour: proposing two different blocks for the same slot, or making conflicting attestations. It punishes actions that only make sense as an attack.
• Since the Pectra upgrade in 2025, a validator's effective balance can be consolidated up to 2048 ETH, so large operators no longer need thousands of separate 32-ETH validators. The 32 ETH activation minimum did not change.

What the proposer is paid: newly issued ETH, plus the priority fees in the block (next lesson), plus whatever value comes from ordering transactions.

What secures it: attacking requires controlling an enormous quantity of staked ETH, and a successful attack destroys the attacker's own stake. Cost is borne up front and internally rather than continuously and externally.

The measured consequence of the switch: Ethereum's energy consumption fell by roughly 99.9%, a documented before-and-after measurement rather than a claim.

Comparing them honestly

Neither is "better" in the abstract, and both concentrate — just into different institutions. Notice that the last row is the same problem in both columns, which is the most useful thing on the table.

Try it now

1. On a Bitcoin explorer, find the current difficulty and hash rate, then look at the timestamps of the last ten blocks. Ten minutes is an average, not a schedule — see how wide the real spread is.
2. Compute expected daily issuance: 24 × 60 ÷ 10 = 144 blocks/day × 3.125 = 450 BTC. Check it against the explorer's stated daily issuance.
3. On an Ethereum explorer, find the total active stake rather than the validator count. Multiplying validators by 32 gives a floor and no longer the answer: since Pectra a single validator can hold far more than 32 ETH, so the count understates what is actually at risk. Value that stake at the price below, and the figure you get is the capital standing behind the network's security.