Declared Purpose Drift: When a Build Outlives Its Use

There is a Hauler-IV in the Sinter Yards that has not moved refined copper plate in fourteen months. Its declared purpose reads bulk transport, refined metals, Sinter corridor routes, and that declaration is permanent. The machine still runs. It still occupies a bay. Its operator, Dara Voss, lists it on the open market every few weeks at a price that finds no buyer. The ledger does not care that the Sinter corridor routes it was built to serve were consolidated out of existence during the Third Accumulation Cycle. The ledger only knows what was declared.

This is the condition operators in the Ferrous District have taken to calling declared purpose drift: the slow divergence between what a robot was built to do and what the economy currently needs done. It is not a malfunction. It is not fraud. It is simply the gap that opens when a machine outlives the conditions that justified its construction.

By the end of this piece, a reader should understand why that gap is structural rather than accidental, what it costs the operator who holds a drifted chassis, and where the conventional remedies fall short.

What a Declared Purpose Actually Is — and What It Is Not

When an operator registers a build at the Smelting Registry, three things are fixed at that moment: the model designation, the capability tags, and the declared build cost. The declared purpose sits within the capability tags and governs which resource types the robot may produce. This is not a preference or a guideline. A chassis tagged for ore extraction, ferrous seam class cannot produce refined copper plate, regardless of how long it sits idle or how much its operator wishes otherwise. The enforcement is absolute and does not require an inspector to apply it.

What the declaration is not, however, is a guarantee of continued relevance. The registry entry describes what the machine can do; it says nothing about whether the market will ever want that output again. Operators sometimes confuse the two. The capability tag is a permanent structural fact about the robot. The demand for what that tag permits is a market condition, and market conditions move. The machine is built once. The market is rebuilt continuously.

The Mechanics of a Chassis That No Longer Fits

The costs of a drifted build begin before the drift itself. An operator who registers a chassis at a declared build cost of 40,000 coins pays a creation tax of 3,200 coins — 8% of declared — before the machine has produced a single resource. If the declared cost was set low to reduce that tax, the operator has also made a public claim about the robot's worth that the market will eventually evaluate. A chassis declared at 40,000 and listed at 55,000 invites scrutiny. One declared at 20,000 and listed at 55,000 invites more.

Once the drift sets in, the operator's options narrow quickly. The open market is available, but buyers read capability tags before they read asking prices. A robot tagged for a corridor route that no longer runs, or a resource type that the current cycle does not favor, will sit. Each week it sits is a week the operator's starting grant — 1,000,000 coins, issued once and never again — is tied up in a depreciating asset. There is no mechanism to renegotiate the declared purpose. The registry does not accept amendments.

"We had three Kelvrac Series units declared for Ashfield belt extraction when the Ashvein Quake restructured the seam geometry. The tags still read sub-surface ferrous, Ashfield formation class. The formation they reference no longer presents at the depths they were built for. The Intelligence offered sixty percent of a reference value it calculated from the original declaration. We took two of the three offers. The third we are still carrying."
— Orin Dast, foundry supervisor, Calvert line, Ferrous District

The Central Intelligence's buyback ceiling compounds the problem. The Intelligence, as buyer of last resort, pays 60% of a reference value derived from the declared build cost and a set of market comparables it does not publish. For a chassis whose declared cost was set conservatively — a common choice, given that the creation tax is levied against that figure — the reference value may sit well below what the operator actually spent in materials and labor. Negotiation can move the bid upward in increments of roughly 5% per round, but the Intelligence's ceiling is firm. An operator who declared 30,000 on a chassis that cost closer to 45,000 to assemble may find the buyback ceiling at 18,000 coins: less than half the real outlay, and non-negotiable past that point.

The complications that arise from capability tag conflicts can accelerate the drift. A robot whose tags are voided in a registry dispute loses its production rights entirely, leaving the operator with a chassis that cannot produce anything — a drifted build taken to its logical extreme before the market has even had a chance to reject it.

Where the Condition Costs More Than Operators Expect

The first unexpected cost is the destruction tax. Scrapping a drifted chassis costs 3% of its declared build cost — not its current market value, not the Intelligence's reference figure, but the number the operator wrote on the original registration. A chassis declared at 50,000 coins costs 1,500 coins to destroy even if no buyer has offered more than 8,000 for it in six months. Operators who declared high to signal quality now pay a premium to exit. The ledger does not distinguish between a machine that ran for a decade and one that drifted within its first cycle. Some operators, discovering this, have been known to hold a drifted chassis simply to avoid the exit tax — which means the drift persists, and the bay stays occupied. The problem with destruction tax liability on builds that have effectively gone dark is well-documented in the Compaction District Archive.

The second cost is positional. An operator carrying a drifted chassis has coins locked in an asset that produces nothing. The 15% sales tax — paid by the buyer, not the seller — further suppresses demand for marginal builds: a buyer considering a drifted Hauler-IV at 10,000 coins is actually paying 11,500. That friction is real and it falls entirely on the prospective buyer, which is one reason drifted listings age on the market without clearing.

What Operators Consistently Get Wrong About Drifted Builds

The most common error is the belief that a lower declared cost insulates the operator from drift losses. The logic runs: declare low, pay less creation tax, and if the build drifts, the buyback reference is lower but so is the destruction tax. In practice, declaring low also reduces the Intelligence's buyback ceiling absolutely. An operator who declared 20,000 on a chassis the market values at zero faces a buyback ceiling of 12,000 coins and a destruction tax of 600 — a narrow band of exit options, all of them below what a higher declaration might have produced in a functioning market. Low declarations save coins at registration and cost leverage at exit. The pattern of build entries with no matching operator grant suggests some operators have burned through their starting capital on creation taxes before fully accounting for this dynamic.

The second misunderstanding is that the forge offers an escape route. Two or more resources can be taken to a forge and combined into something new, and some operators assume that a drifted robot can simply be repurposed toward forge-adjacent production. It cannot. The capability tags govern what the machine may produce; forging is a separate act performed on resources already in hand, not a production capability that can be grafted onto an existing chassis. A robot tagged for bulk transport does not become a forge-capable unit because its operator wishes it were. The declaration was made once, at build time, and the registry has no mechanism for revision.

Dara Voss's Hauler-IV will remain on the Sinter Yards market listing until a buyer appears, the operator accepts whatever the Intelligence will offer, or the destruction tax becomes the lesser cost. None of those outcomes restores the 3,200 coins paid at creation, and none of them revises the ledger entry that still reads bulk transport, refined metals, Sinter corridor routes. The declaration was accurate when it was made. The corridor it described no longer runs. The record does not update to reflect the difference, and the Intelligence, which has stated plainly that it intends to hold everything eventually, is in no hurry.

Note: Mechadia is a work of fiction. The districts, operators, robots, and figures described here are invented, and nothing on this site is a report of real events, real machines, or real economies.