Course

Crypto & Society

The bigger questions crypto raises: its energy use, the ethics of money, privacy and censorship resistance, and where AI and crypto collide.

Intermediate3 units · 16 lessons~36 min

Certificate

Not earned yet · 0 of 3 unit checks passed

Bitcoin (PoW) consumes meaningful electricity — on the order of a mid-sized country — because PoW security is bought with energy; the exact figure is contested and varies with price and hash rate.
Ethereum cut its energy use ~99.95% by moving from PoW to PoS in the 2022 Merge — PoS security is bought with staked capital, not energy.
The "wasted energy" critique vs "energy is the point" debate turns on whether PoW’s energy spend buys something valuable (censorship-resistant settlement) and whether the energy is grid-beneficial or harmful.
Miners follow cheap/stranded energy (excess hydro, flare gas, remote geothermal) — the same kWh is not equivalent across locations, and mining can monetize energy that would otherwise be curtailed.
Carbon accounting is contested: estimates depend on the assumed generation mix; some grids are hydro-dominated (e.g. Quebec, Iceland), others are coal-heavy.
Regulatory pressure (New York’s since-lapsed mining moratorium, EU MiCA environmental disclosure, the EU Taxonomy’s non-eligibility for crypto mining) is forcing transparency on crypto’s footprint. This is educational, not investment advice.

Crypto raises ethical questions because money is infrastructure with distributional consequences — crypto is no exception.
Financial inclusion (Nigeria, Argentina) is a genuine benefit; speculative harm disproportionately hits the least sophisticated users. Both are real.
The PoW energy externality is a policy and ethics question, not just a technical one — see the Energy module.
Privacy serves legitimate uses (dissidents, medical payments) and illicit ones (mixers, sanctions evasion) — the balance is a societal choice, not a purely technical one.
Censorship resistance is crypto’s core political claim: should money be stoppable, and who decides? Stablecoin freezes and OFAC address lists make this real, not theoretical.
"Decentralized" and "trustless" are spectrums, not binaries; the FTX collapse showed that decentralization claims do not protect against operator fraud. This is educational, not financial or legal advice.

AI and crypto intersect in both directions: AI is changing how people trade, attack, and defend crypto — and crypto is building payment rails that AI agents can use.
The biggest near-term user-facing danger is not "AI taking over crypto" — it is the same scams as before, now supercharged by deepfakes and AI-generated content. The classic defenses (out-of-band verification, no guaranteed returns, distrust of urgency) still work.
AI agents need payment rails that do not assume a human with a bank account; stablecoins and protocols like x402 are early attempts, and "who holds the agent’s keys" is a genuinely unsolved custody problem.
Most "AI tokens" are narrative wrappers. Judge them by usage, code, and revenue — not the label. (Educational framing, not financial advice.)
AI data centers and miners now compete for the same powered infrastructure — miners repurposing sites for AI is one of the realest AI × crypto stories so far.
Blockchains cannot natively verify what an AI model did; zkML and oracle designs are early research. Be skeptical of any claim that a chain "runs AI" on-chain.