Blockchain: What It Is and How It Works

What a Block Actually Contains

According to NIST (the National Institute of Standards and Technology, 2018), a blockchain is a distributed digital ledger organized into blocks of cryptographically signed transactions, chained together. The technical definition rests on two properties: tamper resistance and tamper evidence. In practice, anyone who edits the ledger leaves a visible trace for whoever compares copies.

Every block splits into two parts, as NIST explains: a header holding the metadata, and the actual data, meaning the batch of transactions collected in that block. The header does the linking work: it carries the hash digest, the cryptographic fingerprint, of the previous block’s header. Change something inside a block that’s already published, and its hash changes, a change that then ripples forward through every block that comes after, because each one carries the hash of the block before it.

Real-world example: picture a notebook with numbered pages, where the top of each page prints a fingerprint of the page before it. Tear out a page in the middle to rewrite it, and the code at the top changes, which throws off the codes on every page after it. Anyone comparing the notebook against another copy would notice right away.

The practical result is a ledger that’s hard to falsify without leaving a trace, the property NIST calls tamper resistance.

A Network With No Center

The Bitcoin blockchain, per NIST, is independently maintained by a distributed group of participants, with nodes that can sit anywhere in the world. There’s no single server holding the “correct” copy: each node keeps its own ledger and checks it against everyone else’s. It’s an arrangement that echoes how the internet works more broadly, where no central authority is needed to move data from one point of the network to another: blockchain applies that same distributed logic to the problem of keeping a shared, trustworthy record.

The Consensus Problem: Who Gets to Write the Next Block

If no one runs the network, there still has to be a rule for deciding who earns the right to add the next block. Satoshi Nakamoto’s Bitcoin whitepaper, published in 2008, settles this with proof of work: nodes compete to find a nonce, a number used only once, which combined with the block’s data produces a hash starting with a set number of zero bits via SHA-256. Finding it takes many attempts; checking it, once found, is instant. The consensus rule itself is simple: nodes always treat the longest chain, the one backed by the most computational work, as the correct one.

Ethereum ran on the same principle, using the Ethash algorithm, until 2022, when it switched to proof of stake, as Ethereum.org documents. Instead of heavy computation, anyone who wants to become a validator deposits 32 ETH into a dedicated contract and runs three separate pieces of software: an execution client, a consensus client, and a validator client. In every slot, one validator gets picked at random to propose the next block. Act dishonestly, and a validator risks slashing: a penalty ranging from under 0.1% of its deposit, if it’s the only validator involved, up to 100% in a mass-slashing event touching many validators at once.

More Than Payments

Bitcoin’s whitepaper describes a system built to let two parties pay each other directly, with no financial institution in between. NIST lists non-financial uses too: business-to-business transactions that record activity on digital or physical assets, supply-chain traceability (tracking a shipment’s location, for instance) and smart contracts, where a transaction sends data, processes it, and stores the result. These sit apart from cryptocurrency, which remains the best-known use case but far from the only one: blockchain remains the underlying technical structure, and cryptocurrency is just one application built on it.

Seeing that distinction also helps make sense of what an algorithm actually is: proof of work and proof of stake are, in the end, procedures: defined sequences of steps applied to a specific problem, that of getting thousands of computers to agree without a boss.

The Limits, According to the People Who Studied It

NIST is blunt on one point: blockchain isn’t a cure-all, and it comes with trade-offs to weigh case by case. The resource demands of proof-of-work systems, NIST notes, have pushed publishing nodes toward regions with a surplus of cheap electricity. User security isn’t automatic either: an attacker who wants to target people using the network still has to track down each participant individually.