Who actually runs a blockchain, and why would they?
There is no company. No head office, no server room, no CEO who can be phoned. So who is running this thing, and what stops them from cheating?
Two different jobs, constantly confused
Full nodes download the entire chain and independently verify every rule: signatures valid, coins not already spent, block size within limits, subsidy correct. A node that receives an invalid block simply discards it. Running one requires a normal computer and a few hundred gigabytes of disk. It earns nothing.
Block producers — miners on proof-of-work chains, validators on proof-of-stake chains — assemble pending transactions into candidate blocks and compete for the right to add the next one. This is the paid job.
The distinction is the whole security model. Producers propose; nodes decide what is acceptable. A miner who produces a block paying itself twice the allowed subsidy has wasted its effort — every node rejects it, and the block earns nothing. Nobody has to intervene, because everybody is checking.
The community phrase for this is don't trust, verify, and it is meant literally: the software on your machine re-derives the truth rather than asking anyone for it.
Why anybody bothers
Block producers are paid in two ways on Bitcoin, and it is worth knowing that this is the simple case:
- Newly issued coins — the block subsidy, created by the block itself.
- Transaction fees — paid by the people whose transactions are included.
So honesty is not assumed; it is made more profitable than the alternative. A producer that follows the rules gets paid. A producer that breaks them spends real resources on a block the network throws away.
What decentralisation actually costs
Because every full node re-executes everything, throughput is bounded by the slowest participant you want to keep, not the fastest.
- Bitcoin processes roughly 7 transactions per second.
- Ethereum's base layer handles low tens per second.
- A card network handles tens of thousands.
That gap is not a bug awaiting a fix; it is the price of having no middleman. Every design that raises throughput does so by relaxing something — bigger blocks (fewer people can run a node), fewer validators (more concentration), or moving execution off the main chain (a different trust model).
Changing the rules: forks
With no authority, rule changes happen by adoption. Software is proposed; node operators choose whether to run it.
- A soft fork tightens the rules, so old nodes still accept new blocks.
- A hard fork loosens or changes them, so old nodes reject new blocks — and if a meaningful group refuses to upgrade, the chain splits into two, each with its own history up to the split point and its own coin thereafter.
This has happened for real. Ethereum split into Ethereum and Ethereum Classic in 2016 over whether to reverse the effects of a large contract exploit. Bitcoin split into Bitcoin and Bitcoin Cash in 2017 over block size. Governance disputes in this world do not end in a vote; they end in a copy.
Decentralisation is a spectrum, not a badge
Be honest about the picture. On Bitcoin, most hash power is coordinated through a handful of mining pools. On Ethereum, a large share of stake sits with a few staking providers and liquid-staking protocols. Both networks have thousands of independent verifying nodes and a much smaller number of entities that actually produce blocks.
So the correct question is never "is it decentralised?" but "decentralised in which layer, and how concentrated is the other one?"
Try it now
- On a Bitcoin block explorer, look at the last 20 blocks and note the pool name credited with each. Count how many distinct names appear. That count is a rough, live measure of production concentration.
- On an Ethereum explorer, find the total number of active validators. Compare that number to the number of entities that operate most of them — the explorer's staking pages usually break it down.
- The two networks' prices are below. Measure the largest fall you can find on each over the same window and write the two percentages side by side. Two separate networks, two rule sets, two sets of people running them — and two entirely separate price series, both of which have fallen sharply in past periods.