Proof of work is usually explained as computers solving puzzles, which is accurate and unhelpful, because it does not explain why the puzzle matters.

The problem being solved

A distributed system with no central authority needs a way to agree on the order of transactions.

Which is difficult because participants can lie, can be offline, and can rejoin with a different view of history.

The specific difficulty is that identity is cheap online, so a voting system can be captured by creating many identities.

What the work does

It makes influence proportional to something costly rather than to something free.

Which means creating extra identities gains nothing, since the cost is in computation rather than in identity.

That is the actual function — converting a problem of identity into a problem of expenditure.

The hashing

Miners repeatedly hash a block header with varying values, looking for an output below a target.

Which cannot be shortcut, since hash functions are designed so that the only route to a specific output is trying inputs.

Verification is trivial, which is the asymmetry the whole system depends on — hard to produce, easy to check.

Difficulty adjustment

The target adjusts periodically so that blocks arrive at a roughly constant rate regardless of how much computation is applied.

Which means adding hardware does not produce more blocks, it produces a higher difficulty.

This is the feature that keeps issuance predictable while allowing the security budget to grow.

The longest chain rule

Nodes follow the chain with the most accumulated work rather than the longest by block count.

Which resolves disagreements automatically, since a competing chain must out-work the existing one to be adopted.

Rewriting recent history requires redoing the work since the point of divergence while the honest chain continues to extend.

What a majority attack could and could not do

An attacker with a majority of hash power could reorder or exclude recent transactions and double-spend their own.

They could not create coins from nothing, spend coins they do not hold, or change the rules, since nodes validate those independently.

Which is a narrower attack than the popular description suggests, and it is still serious for anyone accepting payments.

Confirmations

Each subsequent block increases the work required to reverse a transaction.

Which is why merchants wait for confirmations proportional to the value at stake rather than applying a fixed number.

The energy question

The expenditure is the security, which is why proposals to make it cheaper generally reduce what it protects.

Whether that trade is worth making is a genuine argument rather than a settled one, and it depends on what the system is for.

This is technical explanation and is not investment advice.

The mining pool structure

Individual miners join pools to receive smoother payouts rather than winning rarely.

Which concentrates block construction in a small number of pool operators even where hardware ownership is distributed.

The distinction matters, since the pool chooses transaction ordering while the miners provide the work.

Protocols allowing individual miners to construct their own block templates address this and adoption has been gradual.

Selfish mining

A theoretical strategy where a miner withholds found blocks to gain an advantage.

Which has been analysed extensively and is argued to be profitable below a majority share under some assumptions.

Whether it occurs in practice is disputed, and the analysis influenced later protocol design regardless.

Hardware evolution

Mining moved from general processors to graphics hardware to purpose-built chips.

Which concentrated production among a few manufacturers and raised the capital barrier substantially.

Some networks deliberately chose functions intended to resist specialised hardware, with mixed success over time.

The security budget question

Miner revenue comes from issuance and from fees, and issuance declines by design on several networks.

Which raises a long-term question about whether fees alone will fund sufficient security.

This is a genuinely open debate rather than a settled matter, and reasonable people disagree about the magnitude of the concern.

Geographic distribution

Mining follows cheap electricity, which has concentrated it in specific regions at different times.

Which makes the network sensitive to policy changes in those places, as demonstrated when a major jurisdiction banned the activity and hash rate relocated over months.

The relocation itself demonstrated the system continuing to function through a substantial disruption.

What a node actually validates

Signatures, that inputs exist and are unspent, that consensus rules are followed and that the proof of work meets the target.

Which means running a node gives independent verification rather than trust in miners.

Miners cannot change rules that nodes enforce, which is the check on their power.

Why the design is hard to change

Consensus rules can only change if operators adopt the change, which makes the system conservative by construction.

Which frustrates people who want faster development and is the property that makes the rules credible.

A monetary rule that can be changed easily is not a rule in the sense the design intends.