The facade Grigory Potemkin is said to have raised along the Dnieper for Catherine II’s 1787 tour — and its long Soviet descendant, приписки (pripiski, padded plan reports). The centre is genuinely runnable: it makes the gap between the reported and the real mechanical, watches the hidden shortfall accumulate, and measures the one hard fact — a report that always says “met” carries zero bits.
source The 1787 Crimean tour and the “Potemkin village” legend (a facade to impress a sovereign); Soviet pripiski — falsified production reports under plan pressure; C. E. Shannon, “A Mathematical Theory of Communication” (1948). Room THE STATE. Figures computed live; nothing quoted.
In 1787 Catherine II toured newly annexed Crimea. The legend says her governor Grigory Potemkin ran up painted facades of prosperous villages along the route — a front with nothing behind it. Whether or not he did, the phrase stuck: a Potemkin village is a false front built to be seen, not lived in.
Its Soviet heir is приписки (pripiski): under pressure to “meet the plan,” managers padded their production reports. A paper economy grew alongside the real one — and only the paper one hit its targets. AMBER historical framing.
A report is a channel. Its information is H(p) = −p·log₂p − (1−p)·log₂(1−p), in bits. Slide p = the rate at which the plan is truly met, feeding an honest channel that mirrors reality:
A channel that always says “met” outputs one symbol forever — entropy 0 bits, at every p. It cannot tell success from failure. The honest channel peaks at 1 bit when p=0.5.
Pripiski did not just embarrass — they corrupted the numbers the state ran on. Padded reports flowed upward into planning statistics that were then quietly unreliable, so the centre planned against a fiction.
This is exactly the data problem that THE OGAS and GOBERNET were meant to solve: a real-time loop is worthless if its inputs are padded. Honest measurement is the precondition for control — not an afterthought. Ties both.
Each period the plant makes true units against a plan target. Under pressure it reports R = min(target, true + padding) — padding is the lie added on paper, capped at the target (you cannot over-report past the plan without inviting a raised quota).
While true < target and padding covers the shortfall, the report reads exactly target every period — a perfect record. But the hidden gap R − true = min(target−true, padding) is real, and it accumulates.
—
THE OCHO TIE. Set padding to 0 and the gap is zero, the shortfall stands in plain sight, and the channel can finally report failure. That is the whole discipline of this OCHO: every LIT instrument keeps a fail-loud that must make noise — because a report that can never say “no” carries no signal at all. LIT
The facade does not remove the shortfall; it only delays and concentrates it. Reality is the auditor that cannot be padded.
“The plan was overfulfilled.” Cut. A report pinned to “met/overfulfilled” regardless of output is a facade — a constant-success signal is information-free (0 bits, centre).
“If every plant reports success, the economy is healthy.” Examined. Uniform success is precisely the zero-entropy case: the channel has lost the ability to distinguish a healthy plant from a failing one.
“Padding is harmless bookkeeping.” Cut. The hidden gap is conserved and accumulates every period; the goods are simply missing, and the shelves prove it.
Claim the always-“met” report is informative and trustworthy — assert it carries a full bit of signal. The honest witness measures the channel’s actual output entropy, finds 0 bits, and the mismatch turns the witness red.
A channel that cannot report failure conveys nothing. Claiming otherwise does not add a bit — the measured entropy is 0, and the witness disagrees.