Circles gives every person their own personal currency, continuously issued, made spendable through webs of trust: you accept someone's Circles only if a chain of trust connects you to them. The Alignment Economy issues one identical currency to everyone and handles trust at the account level instead. The two projects share more DNA than any other pair in this space, and the one difference between them decides whether the money is actually money.
Where they agree
Circles is the closest existing relative to the Alignment Economy, and it got the radical parts right first. Equal ongoing issuance to every person, simply for being a person: Circles shipped that in 2020. Built-in decay on balances so the money circulates rather than accumulates: Circles has that too. Sybil resistance through social connection rather than government ID: same instinct. On issuance philosophy, the projects are siblings.
The comparison
| Dimension | Circles | Alignment Economy |
|---|---|---|
| The unit | Each person mints their own personal currency | One global currency, identical points for everyone |
| What a unit is worth | Depends on who issued it and who is looking: value flows through trust paths | Every point is worth every other point, everywhere |
| Where trust lives | In the points (your currency is only good along your trust graph) | In the account: a percent-human score multiplies purchasing power, points stay uniform |
| Sybil resistance | Trust connections limit how far a fake person's currency travels | Miner verification, staked vouching, courts, bounties |
| Pricing a loaf of bread | Depends which Circles the buyer holds and whether the baker trusts their issuers | 20 points, for everyone |
| Anti-hoarding | Demurrage on balances | Daily expiry of allocations plus a share-preserving rebase on savings |
The fungibility failure
Money works because it erases history. A dollar is a dollar whether it came from a saint or a scoundrel, which is exactly what lets a stranger sell to a stranger. Economists call this fungibility, and it is not a nice-to-have; it is the feature that makes prices possible.
Personal currencies break it on purpose, hoping trust graphs can route around the damage. In practice the damage wins. A shopkeeper in a Circles economy cannot post a price, because "50 Circles" is not one quantity; it is a different quantity for every customer depending on whose Circles they carry and how the trust paths run. Every transaction becomes a small foreign-exchange negotiation. Liquidity pools and hub currencies were built to patch this, and each patch quietly reintroduced the uniform currency that the design had rejected.
None of this is a failure of execution. Circles has been iterated by capable people for years. It is a failure baked into the premise: observer-relative money cannot host an economy of strangers, and an economy of only friends does not need money.
The lesson the Alignment Economy took
Score the account, not the points. The problem Circles used personal currencies to solve (how do we know an account is a real, unique human?) is real, but the answer belongs on the account, not in the unit. In the Alignment Economy, an account's percent-human score discounts its purchasing power at the moment of spending, and then the points move on, indistinguishable from all others. Trust is enforced; fungibility survives; the baker posts one price.
Same diagnosis as Circles. Same egalitarian issuance. The trust layer moved eighteen inches, from the money to the account, and that move is the difference between a social experiment and a currency.
FAQ
Is Circles dead? No, it continues to evolve, and its redesigns keep converging toward more uniformity, which rather proves the point.
Doesn't the percent-human discount also break fungibility? No. The discount applies to the spender at checkout, like a fee. The points received by the seller are full, ordinary points with no memory of who spent them.
Why compare to Circles at all if it's small? Because it is the only other project that got issuance right, which makes it the cleanest controlled experiment on the one variable that differs.