Step 6 / The Bitcoin Machine
What the work protects.
Explain what computing work makes harder, and which problems it does not solve.
Maya’s payment is in the shared history. Replacing that history takes more than editing a file. An attacker must compete with work that other miners have already done and may still be adding.
That work uses real equipment and electricity. It has a particular job in Bitcoin’s design: making a competing accepted history costly to build under the network’s assumptions. Spending energy is not a guarantee that every payment is honest or every attack fails.
The replacement still has to follow the rules
Before comparing work, a full node checks whether a block follows its rules. Wrong signatures, inputs that are not available, and excessive rewards all cause rejection. Doing more work does not remove those checks.
Now suppose an attacker originally paid a merchant. The attacker tries to replace that recent payment with a different payment using the same old output. In the competing history, the attacker spends that original output differently.
The attacker is trying to change which payment becomes part of the lasting record. They are not learning the merchant’s private key. The target is the order and persistence of history, not the ability to sign with any key they choose.
The upper branch has more illustrated work, but violates the reward rule. It is rejected.
Mining power gives some powers, not every power
A miner chooses which transactions to put in its own proposed blocks. It may keep leaving out one user’s valid transaction. Deliberately blocking a transaction this way is censorship. The transaction can still follow the rules even though the miner refuses to include it.
Enough sustained hashing power can support an attempt to replace recent accepted history. Its chance of success and its cost depend on the circumstances. The attacker is competing with other miners and the work added to the history.
This power does not let the attacker sign arbitrary payments with other people’s keys. It also does not force nodes using unchanged rules to accept too much new bitcoin. Keep the power to influence inclusion separate from the rules that make a payment valid.
The same electricity use can mean different emissions
Imagine two mining operations use the same amount of electricity. One gets it from low-emission generation and the other from high-emission generation. Their electricity totals match, but the greenhouse gases linked to their use can differ.
Those released gases are called emissions. To estimate mining’s environmental effect, we need to know where and how its electricity was made. We also need to ask what would have happened to that electricity or fuel if mining had not used it.
A low power price does not prove low emissions. One useful local arrangement does not show that every mine has the same effect. Look at what an estimate measures, what it leaves out, and how it was calculated.
The resources behind mining can change
Miners receive the allowed new subsidy and the fees in their blocks. They use income to help cover equipment, electricity, and other costs. If the subsidy falls, one part of that income changes.
That is not the whole calculation. Fees may change, bitcoin’s price may change, and equipment or energy may cost more or less. Miners can respond to those changes. An attacker’s resources also matter when judging the risk.
So halving the subsidy does not mechanically halve security. It also does not prove the future will take care of itself. We need to examine the incentives and resources, rather than turn one revenue change into a price prediction or a guarantee.
Another way to put something at risk
Ethereum uses a different system called proof of stake. Participants commit ETH, run software, and perform duties. These participants are called validators. A selected validator proposes a block, while others sign statements about the chain.
These signed statements are called attestations. Validators can earn rewards and face penalties. Certain rule violations can cause some of their staked ETH to be taken away, a penalty called slashing. Making empty usernames does not create more ETH to stake.
The transactions still have to follow the execution rules: the rules for payments and programs. Having a large stake does not make a bad signature valid. This is the same distinction between influence and validity we saw in Bitcoin.
Ethereum also uses agreement weighted by stake at checkpoints to finalize history. Finalized history is treated as settled under the agreement rules. Undoing it would require a serious failure of those rules, with large penalties for validators. First appearing in a block is an earlier step. This is different from Bitcoin’s growing proof of work.
The payment records differ too. Ordinary ETH transfers update account balances; they do not use Bitcoin’s UTXO model. A smart-contract transaction can also change a program’s stored records.
When comparing networks, keep asking what the user approved, which rules were checked, and what changed. Learning one blockchain gives you questions for the next. It does not mean every coin works the same way.
The idea to keep
Ask what an attacker must do, what it would cost, and what honest participants check for themselves. Then name the limits. Using energy or stake alone is not a full explanation of security.
Next, separate enforcing today’s rules from persuading people to adopt new ones. Those are different kinds of control.