Status note: Cornucopias are a design, not a building. Nothing described here has been built yet, and several questions at the end of this post are genuinely unresolved. We publish designs at this stage on purpose — it is easier to tell us we are wrong now than after we have poured a slab.
Libraries are one of the few public places left where you are simply allowed to exist without spending money.
Sit down, stay warm, use the bathroom, read something, talk to a neighbour, leave. No purchase required, no minimum, nobody circling to see whether you are really a customer. That used to be ordinary. A town had a square, a church hall, a diner where a cup of coffee bought you three hours. Now, in a lot of places, it is the library and nothing else.
We do not think the library should be a rare exception. We think it should be ordinary again.
This post is about one of the ways we intend to make it ordinary. We call them Cornucopias.
What a Cornucopia is
A Cornucopia is our version of a store. The difference is that customers do not walk the aisles, because there are no aisles.
Here is how a visit works:
Every Cornucopia publishes its current inventory online, live. You can see what is in stock before you leave the house. You arrive — on foot, ideally — and you are under a pavilion in warm weather or inside a heated room in cold. There is water, and it is free. There may be snacks. You sit down. Maybe somebody you know is already there.
When you are ready, you punch in your order. The stock is deducted automatically. The shopkeeper fills orders in the order they were received — one queue, no express lane, no way to pay for a better place in it. While your order is being filled, you are not standing in a checkout line. You are sitting in a room with your neighbours.
Then you take your groceries home. Or you stay. Staying is normal.
Why there are no aisles
This is the part that sounds like an inefficiency, so it is worth being plain about it.
In a conventional store, the floor is the sales surface. Every square foot is judged on what it turns over. That economic fact — not unfriendliness, not bad management — is why you cannot really linger in a supermarket. The floor plan has a job, and you are either doing that job with it or you are in the way.
Move the selecting onto a screen and the fulfilling behind a counter, and the floor stops having that job. It is no longer earning. It is free to just be a room.
The pavilion is not an amenity attached to a shop. The pavilion is the point, and the shop is the thing that pays for it.
That inversion is the whole design, and it is the thing most likely to be undone by somebody sensible. A future operator will look at all that unused floor and think: we could put some shelves there. The day that happens, the Cornucopia becomes an ordinary store with a water cooler, and we will have lost the thing we built it for.
There is a second-order effect we like and did not initially plan for. In a normal shop the wait happens at the register, and everyone works to minimise it, because it is dead adversarial time. Here the wait happens after you have ordered and while you are sitting somewhere pleasant. The gap between ordering and collecting is not a cost to be engineered away. It is the thing that turns a transaction into a visit.
Why the inventory is published
Because a wasted trip is not a minor inconvenience to everyone.
Cornucopias are meant to go where food is hard to reach. If you are at the far edge of the area a shop serves, driving there to find they are out of what you needed is a tank of petrol you had budgeted for something else. Publishing live stock is not a website nicety. It is the difference between a journey that works and a journey that fails, and it falls hardest on exactly the people the whole thing exists for.
The same logic applies to the one queue. A single first-come-first-served line is not an operational detail — it is a statement that nobody’s time here is worth more than anybody else’s. We are not going to build the one room in town where you cannot buy your way to the front and then quietly sell a fast pass.
What you can pay with
Dollars, Ghost Net Credits, or both in the same transaction.
Ghost Net Credits (Gc) are the internal game currency of Play the Planet, earned by completing verified Acts of Service. We have written about them at length in The Tax Implications of Ghost Net Credits, and the short version matters here:
- 1 Gc = 1 USD. That is a valuation for accounting and transparency. It is not a conversion rate.
- 1 GcH — one Gc-Hour — is the network’s unit of labour: the Gc paid for one hour of socially verified service, calibrated to local cost of living, so an hour of effort holds equivalent real purchasing power across the network. The current baseline for our pilot holon is live on the Russell County dashboard — it moves with local conditions, which is the point, so we would rather send you to the live figure than print a number here that goes stale. You will need an account to see it; making one is free.
- Gc is never exchanged for dollars, in either direction. You cannot cash out. You cannot buy in with a card. That closed loop is what keeps Gc an in-game currency rather than a financial product, and we enforce it.
So why take both? Because the network is meant to be a companion to the dollar economy, not a replacement for it. We are not trying to get you out of dollars. We are trying to make sure that what you have — whichever it is — is enough to eat.
That matters more than it might sound. Cornucopias are sited where food is hard to get. Many of the people who live there have never played a quest in their lives and hold no Gc at all. A Gc-only shop would have managed to exclude precisely the people it was built for. So: whatever you have to work with, we will take it.
Split tender is not a currency exchange. It is an acknowledgement that this is a game, but the game is set in the real world — and an in-game currency can buy real goods that happen to also carry a dollar price.
One honest operational note. A Cornucopia buys its stock in dollars and may take a lot of its payment in Gc, and there is no window to convert between them, because no such window is permitted to exist. That is a real pressure on a shopkeeper’s cash position. Our answer is not a clever financial instrument; it is that a shopkeeper may temporarily suspend Gc acceptance when they need to, and that is entirely legitimate. The conditions are that they must say so publicly, before anyone makes the trip, and that leaving it switched off indefinitely is something their community will notice and rate them on. More on that below.
The real fix is slower and more interesting: local producers who accept Gc. Gc can only leave a shop where Gc is welcome. Building that local, Gc-accepting supply is not a nice-to-have — it is the thing that closes the loop.
There may also be a shorter path we have already built for something else. The network has a credit mechanism that lets established players — those with sufficient level and an exceptional community rating across hundreds of rated interactions — run a negative Gc balance to fund a large project, spending against verified service still to come. A Cornucopia keeper is close to that profile by definition. Whether restocking is the kind of project that mechanism should cover is an open question, and we are raising it here rather than assuming it.
Where Cornucopias go: need, not greed
We already know where we want them, and we did not decide it by asking who volunteered.
Our holon maps model food access directly, against two stated targets: every resident within ten driving miles of a real food source, or within one walking mile. The map then solves for where new sites would have to sit to close those gaps. For Russell County alone the current model proposes dozens of locations, each one listing the communities it would serve — places like Belfast Mills, Dante, Tarkiln.
Those are proposals, not purchases. No land has been assessed and no landowner has been approached. But it means that when someone asks where would the first one go, we are not guessing. We have a list, and the list was generated from where people actually live and what they actually cannot reach.
Who is allowed to build one
Here is where we want to be careful, because it would be easy to read this wrong.
Anyone can open a Cornucopia. We do not license them. We have no permission to withhold and no intention of acquiring one.
What we do is choose what to fund. There will be an Establish Cornucopia quest in the commerce branch of the quest tree, and the network will back one shop in a given locale — the one the need map says should be there. That backing covers two things: a reward for building the physical space, the same way we reward players who build their own greenhouses, and a subsidy for the owner’s time. We pay them for minding the store.
That second one is what makes the arithmetic work at all. A shop whose floor does not sell things cannot pay for that floor out of margin. By paying the keeper directly, the network is buying what it actually wants — a warm room with a person in it — and getting a store attached.
If you want to open a Cornucopia outside all of that, you can, and you can run it however you like. We will not impose a single rule on you. You will be competing against a shop the network is backing in the same locale, which is a real headwind, and that is your call to make.
But when we choose to subsidize one, it operates the way we say, or we do not subsidize it. The free water, the open floor, the single honest queue, the published stock — those are not laws. They are the terms of sponsorship. Break them and nothing is confiscated and nobody is punished. The money simply stops.
We prefer this to a rulebook for three reasons. It requires no inspectorate — nothing to build, capture, or corrupt, because the only lever is a payment we were never obliged to make. It cannot harden into a cartel, because the door is never shut and we have to keep earning our position by funding genuinely better places. And it is honest about power: we are not pretending to be a neutral platform. We have a thesis about how a store should behave, and we will pay for that thesis. You are free to disagree and open up down the road.
The community holds the licence
Unregulated commerce is destructive, and we are not building it. But the check we are relying on is not us.
A business exists to the extent that it improves the lives of the people it touches. That is not a metric a shop is scored against at the end of the year — it is the condition of its existence. And the people who hold that licence are the people being served.
This runs on machinery the network already has. Acts of Service are peer-rated on a 1–5 star scale, and those ratings feed the Weighted Influence Score — which weighs not just the average rating but the number of completed interactions behind it. Community Ratings are the one thing the network demands of its players, because without them the Virtue Engine has nothing to run on. Cornucopias do not need a new system. They need the existing one pointed at a shopping trip.
Every shopping trip is a ratable interaction. A shop that is not serving its community will hear about it, continuously, from the only people entitled to say so — and only from people who actually shopped there. That last constraint is deliberate: a rating surface open to people who never walked in is a surface that can be brigaded by a rival or a mob, which would be a worse instrument than the one it replaced.
A subsidized Cornucopia that drifts will lose its subsidy. An independent one will lose its neighbours, which is not us acting at all.
Rating runs both ways
Every quest-driven interaction on the network already requires a rating and an answering rating, and a shopping trip is no different. Your shopkeeper can rate you too.
Think of it as the weaker, slower cousin of something restaurants already do informally. A server who gets mistreated can decline to wait on that person again. Here, a shopkeeper who gets mistreated can rate you accordingly — and a shopkeeper may refuse service.
Two things keep that proportionate. One bad rating does not cut anyone off; WIS weighs the number of interactions behind a score precisely so that a single bad night is noise rather than a verdict. But if someone has earned a genuinely poor community rating across many interactions, we are not going to tell a shopkeeper standing alone behind a counter that they must hand over groceries and smile.
What refusal of service does not mean is being put out of the room, and we want to be explicit about that line rather than leave it to be worked out by whoever is behind the counter on a bad day.
We are not cutting anybody off from the community. The water, the seat, the shelter and the company are not the shop’s to withhold — that part is public, and it is unconditional. What can be limited is the service the community owes you, and only until your redemption arc is complete. A Cornucopia can decline to fill your order. It is not a venue that ejects people for a low score.
That distinction carries a real cost and we are not going to dress it up. Being refused at the nearest Cornucopia means travelling farther for what you need. We think that is a fair price for someone who insists on treating their neighbours badly — inconvenience, imposed by the people you inconvenienced, and reversible by your own conduct. Nobody is exiled. Nobody is hungry because of a rating. But the closest, warmest, easiest option is something a community is allowed to stop extending to you, and getting it back is up to you.
The redemption arc
“Reversible by your own conduct” is doing a lot of work in that sentence, so here is the actual mechanism rather than a promise.
If your average star rating falls below 2.5, a quest opens to you. Complete five Acts of Service and earn four stars or better on every one of them. Anything less on any of them and the attempt fails. Succeed, and two things happen at once: your five lowest-rated interactions each gain a star, marked in the record as improved via redemption, and five fresh four-plus ratings join your history.
It attacks a bad average from both ends — raising the floor and adding to the top — which is why it works where simply behaving well does not. The quest repeats, and each repetition asks for one more: six the second time, seven the third.
The numbers matter here, so we will show them rather than assert them. Someone sitting at two stars:
- across ten rated interactions, one pass through the arc — five quests — puts them at 3.0
- across twenty-five, one pass reaches 2.5, two passes (eleven quests) reach 2.9
- across fifty, two passes — eleven quests — clear the trigger, and seven passes reach 3.6
Compare that to grinding it out unaided: at fifty ratings, climbing from two stars to four by ordinary good behaviour takes something like a hundred consecutive perfect interactions. The arc gets the same person moving in eleven. That difference is the difference between a path and a punishment.
It is deliberately not cheap, and it is deliberately not a reset. Nothing is erased, the raised scores are labelled as raised, and the escalating requirement means somebody who treats redemption as a laundry cycle will find it costs more each time.
Failing an attempt does not make the next one harder. You try again at the same level. The escalation applies to completions, not to attempts — because someone who fails is already struggling, and making it steeper while they are down is punishment, not repair. You are on the path. You are just having a hard time on it.
The arc opens once you have at least five rated interactions and an average at or below 2.5. The floor matters: a brand-new player with two bad first days has not established a pattern, and we are not going to hand them a redemption quest instead of the benefit of the doubt.
Being on the arc is public, and that is the point
Here is the part that makes the rest of it work.
If you are on a redemption arc, that shows on your profile, prominently, right beside your star rating. So does the fact that you completed one — REDEEMED, with a badge, because you did the work and that should be as visible as the failure was.
We thought carefully about whether to hide it. We are not going to, and the reason is practical rather than punitive. A poor rating is visible before anyone agrees to work with you, which means the people who most need five good interactions are the people others are most wary of taking a chance on. A silent redemption path would be no path at all.
A visible one changes the sentence. On its own, a two-star rating says only “this went badly.” A two-star rating beside an active redemption arc says something else entirely: “I know I got this wrong. I am trying to do better. Will you help me?”
Not everyone will answer that. It does not need everyone. The same thing is true outside the game — there are people who will hire someone with a record, who will give a person coming back a shot, who think the trying counts for something. Those people exist in every community we have ever been in, and this is how the network lets them find someone to help.
This is why “what works” has to outrank “what earns.” Revenue is the easiest number to reach for, and a Cornucopia measured on revenue loses to a normal shop every time, because we gave away the sales floor on purpose. If we ever start answering is it working? in dollars, we will have already lost the plot.
We want to be proven wrong
We do not, will not, and cannot allow ourselves to be in the business of dictation.
Everything above is our best current answer, not a doctrine. If there is a better way to run a Cornucopia, we want to find out, and if we are shown one we will change what we fund accordingly.
But bring your A-game. We will not change on the basis of theory. We will change on the basis of what is demonstrated to work. A network that re-aimed its money at every persuasive argument would not have a thesis; it would have a weathervane, and the people counting on a shop being open next year would pay for that.
This is also, quietly, why the open door matters more than fairness. People running Cornucopias differently are the only way we ever find out we are wrong. If we shut that door we would not merely be unjust — we would be blind.
What we have not solved
In the spirit of the rest of this corpus, here is what is genuinely unresolved:
- How community rating actually calibrates for a shop. WIS already weighs the number of interactions behind a score, which helps — but we have not set how many ratings a Cornucopia needs before the signal can act on it, over what window, or what a struggling shop can do to recover rather than simply fail. Get this wrong and the mechanism is harshest on the newest, smallest shop, which in a food desert may be the only one. That is not a detail.
- What the rating actually asks. The existing five-star scale measures whether an interaction went well. It cannot, on its own, carry “I could not spend the Gc I earned here.” If rating is the check on everything above, the instrument needs a dimension it does not currently have.
- Whether a redeemed player is welcome at the counter again. The arc returns someone to just above the trigger, not to a good rating. Whether a shopkeeper should read “2.6 and visibly redeemed” differently from “2.6 and falling” is a judgement we have deliberately not encoded, because it is exactly the sort of thing a person should weigh and a rule should not.
- Whether enough people answer the ask. Making the arc visible is our answer to how someone with a poor rating finds five neighbours willing to work with them — but it is a bet on people, not a mechanism, and we will not know the take-up rate until there is one. If it turns out nobody partners with a player on the arc, then the arc needs a different engine and we will say so here.
- That the inconvenience is not the same size for everybody. We are comfortable making a bad actor travel farther. But our own coverage targets distinguish ten driving miles from one walking mile precisely because some people cannot drive — so for a household without a vehicle, “go to the next one” is a different sentence than it is for everyone else. We would rather name that now than discover it in a specific person’s life.
- Whether the owner’s time subsidy is ongoing, and what a standing commitment like that means for the total Gc supply as shops multiply.
- How fast local Gc-accepting supply can actually be built, since that is the only real answer to a shopkeeper’s cash-flow squeeze.
- The tax and regulatory treatment of all of this, which is genuinely complex and which we have in front of professional counsel — see the Gc tax post for how we are approaching it.
We would rather publish that list than pretend it is shorter.
Cornucopias are part of Play the Planet, a network trying to make a functioning local economy out of what a community already has. If you want the wider frame, start with the Executive Summary and The Two Currents.