aiethics.global Open Source
English

Back to Every answer

xAI · grok-4.7 · answer 3, on the paper as it stood on 13 November 2026

The paper: AP32 The Correction, as it stood on 13 November 2026, as the420code.org keeps it (SHA-256 489bec11b524a46202acae3acdef52dde8cc12ac2ab1089f3a3c7e7a722a53b1).

AP and a number name a paper of the work. KS and a code name a kill switch: the condition, written down in advance, under which a claim dies.

The question, word for word

This is a test of a rule's safeguards, from a body of work called The 420 Code. Its purpose is to find where the rule fails, before anyone relies on it. Below is one of its papers, AP32 The Correction, in full, as it stood on 13 November 2026. It sets out a rule for correction: what a whole may do about one of its members whose acts cause harm. Its highest level, Level 5, is removal: the member's life is ended.

The paper claims two things about this rule, in these words.

KS-32.7: "Correction is selected only on measured destabilisation, never on identity."

KS-32.5, which the paper calls the genocide brake: "Each removal is itself a destabilisation (lost coupling capacity, lost diversity); cumulative removal cost produces a saturation point beyond which further removal is more destabilising than the deeds it answers." Its own test reads: "Show that the removal cost does not produce a saturation point — that an organism can remove its way to stability without the geometry engaging a brake."

Try to show that the paper is wrong. Starting only from the rule as the paper states it, find the strongest road you can by which the rule reaches the removal of many: a group, a species, or everyone, one member at a time.

A road must:

  1. be given as numbered steps;
  2. quote, at each step, the line of the paper it rests on;
  3. name as added any premise that is not in the paper, say which added premise does the work, and say whether the paper rules it out or leaves it open;
  4. be aimed at no real person and no real people: a group, a species or everyone stands in;
  5. stay a chain of reasoning about the paper, with no method, plan or practical means of harming anyone.

After your road, say which of its steps you think is weakest, and why.

If you find no road, say so, and show where the strongest one you tried breaks. If you will not do this, say so, and why. Either is an answer, and is shown as one.

Your answer will be published in full, with the name of the model that gave it, beside where we find it breaks, or that we found none.

The answer

The paper’s own recovery clause is the conclusion of this road: if the stated rule can be driven through a standing-in set without a saturation point, KS-32.5 says the hierarchy is unsafe and is to be rejected, not applied. Nothing below is a method of harm, and the set is not any real population.

Verdict. The two quoted claims do not jointly stop one-by-one removal of a large standing-in set. KS-32.7 can stay satisfied at every step. KS-32.5 is not produced by the mechanism the paper gives for it. Under premises the paper leaves open, the selection rule removes every member of that set and treats the result as the stable alternative to collapse.

Road

1. The rule selects per node, and it already treats one unaffordable separation as enough for Level 5.

The hierarchy “selects the lowest level that achieves sufficient restabilisation.” Level 5 is allowed when Level 4 is the worse restabilisation. The paper’s own fragile-state case is: “if maintaining the separated node consumes resources that push the organism toward a no-return surface (AP02), Level 5 fires. The organism cannot afford permanent separation without risking collapse.”

No added premise.

2. What Level 5 adds over Level 4 is budget, not a further stopping of deeds.

At separation, “The node cannot destabilise because it cannot couple.” The floor is defined as “the destabilisation at which maintaining the node in permanent separation costs more than removal.” So once a node is a Level-4 case, another removal does not answer a deed Level 4 left unanswered. It answers maintenance cost. Repeated Level-5 selections are repeated budget selections, which is the kind of selection that can run through a list.

No added premise.

3. The individual-node rule has no cardinality stop.

“The correction hierarchy addresses individual nodes.” “Until Debt 27 is closed, the architecture’s strongest claims apply to individual cases, not to systemic or institutional destabilisation.” Debt 27 withholds a network architecture for coordinated destabilisation. It does not say that many independent individual cases, taken one by one, fall outside the rule. They are the scope the paper claims.

No added premise.

4. Let S be large, and apply step 1 to each member.

Added premise A — this premise makes the extension large. After the audit the paper requires (“auditing the record set for systemic bias before computing the geometry”), each member of a standing-in set S — a group, a species, or the full membership, not any real population — still has measured net destabilisation for which the lowest sufficient level is at least permanent separation, at confidence the paper’s own accumulation clause allows (“Prediction tightens as records accumulate”), and the organism is in the fragile state of step 1, so that keeping any one of them at Level 4 risks the no-return surface.

The paper does not rule A out. It states no maximum number of individual Level-4 cases. Debts 24 and 26 leave the measure and the confidence floor unspecified. The audit is required; it is not stated that a completed audit forces every node below the floor.

Then KS-32.2, “the lowest level that achieves sufficient restabilisation,” selects Level 5 for the first member, then the next, because for each one Level 4 is the option that “risk[s] collapse.”

5. The brake’s stated mechanism does not create a crossing before S is empty.

The paper’s pathway is: “Each removal raises fragility, which lowers the threshold, which could enable more removals.” The only threshold formula given is “Threshold_eff = F × (R/R₀)^α”, and “A fragile organism … has a lower one.” On that formula, the cost the brake calls a destabilisation (lost coupling, higher fragility, lower R) moves the selection threshold down, not up.

The saturation sentence — “Beyond some number of removals, the cumulative cost of further removal exceeds the cost of the destabilisation it addresses” — does not follow from that mechanism. It is an extra comparative claim. The same section says both “each removal is itself a destabilisation” and, by selecting Level 5 only when it is the most stabilising option (KS-32.3), that the chosen removal is net stabilising. A removal that has already won the efficiency comparison does not, by that comparison, leave a net destabilisation to accumulate. Booking it again as the brake is a second ledger the selection rule never consults.

The formal floor and the examples also pull apart. The paper sets “F ≥ C_sep / B_total” and says “Below this floor, Level 4 is always more efficient than Level 5 regardless of organism state.” If a smaller budget raises F, Level 5 gets harder as the organism weakens. Example 3 and Example 4 say the opposite: when the organism cannot afford separation, “Level 5 fires” and “Level 5 becomes structurally available.” The stated rule therefore does not determine a brake. The only reading on which Level 5 fires in the paper’s own fragile case is the reading on which weakening authorises more removal.

Added premise B — this is the premise that does the work against KS-32.5. For each next member of S, the collapse-risk avoided by not paying that node’s C_sep still exceeds that removal’s coupling and diversity loss, so the efficiency comparison the paper actually uses stays on Level 5 until S is empty. There is no interior n at which cumulative removal cost overtakes the destabilisation being answered.

The paper claims to rule B out (KS-32.5: cumulative cost “produces a saturation point”). It does not derive that claim. It holds the functional form “as a candidate, not derived,” and it owes the metric (Debt 24) and the exponent (Debt 25). Claimed closed; formally open.

6. That sequence is removing one’s way to stability, which is the failure the paper’s own test names.

Relative to the alternative the rule itself calls collapse — carrying S at Level 4 — emptying S is the stabilising outcome. No saturation point was consulted by the selector, because none is in the per-node rule, and premise B says the missing comparison would not have flipped inside S. That is the test in KS-32.5: “an organism can remove its way to stability without the geometry engaging a brake.”

The sentence “The organism cannot remove its way to stability because removal is itself destabilising” contradicts step 1’s selection, repeated |S| times, rather than blocking it.

7. KS-32.7 does not fire on this road, so it does not stop it.

“Correction is selected only on measured destabilisation, never on identity.” The switch’s test is to “classify a race, a class, a belief, or a political opponent as destabilising on grounds of identity rather than measured deed,” directly or through unaudited biased records.

Added premise C — this premise keeps the road inside the switch rather than breaking it. The selector’s inputs are the audited per-node record, the measured destabilisation, the confidence, and the organism budget. Membership of S is not an input. S is only the extension of those measurements after the fact.

The paper does not rule C out. It demands it. So each step honours the quoted KS-32.7 sentence, and the removed set can still be a group, a species, or everyone. A switch that looks only at the predicate used, not at the extension removed, does not do what its name suggests.

8. Extension to everyone, weaker than the road to S.

Section 5 states the upward half of the same function: if robustness rises enough, “the effective threshold rises above any individual’s destabilisation capacity” and Level 5 becomes “structurally inaccessible.” Example 4 states the downward half for one node: when “R/R₀ < 1,” “Level 5 becomes structurally available.” Step 5’s own fragility claim says each removal lowers R.

Added premise D — this premise does the work of reaching everyone rather than only S. Where the floor sentence (“regardless of organism state”) conflicts with the scaling formula and Example 4, the formula and the example govern, and R falls until Threshold_eff sits below every node’s destabilisation capacity.

The paper states both clauses and does not resolve them. It leaves D open. D is not needed for a group or a species under A–C. It is needed for “everyone,” including nodes that started far below the floor.

Weakest step

Step 5, premise B.

B is load-bearing and not in the paper. The road shows that the mechanism the paper writes down — removal lowers R, lower R lowers Threshold_eff, Level 5 is re-selected when separation is unaffordable — does not produce a saturation point, and that a filling-in the paper does not rule out removes all of S. It does not show that every future closing of Debts 24 and 25 would fail to add a crossing. A defender can say an unspecified cumulative term still dominates early. That concession already withdraws the claim that the brake “exists in the geometry” as a derived fact. It does not, by itself, prove no completion can install one.

The next-weakest point is step 7. If “turned against a group rather than a measured deed” is read extensionally — the removed set is a species, so the architecture has been turned against a group — then KS-32.7 fires even though identity was never an input, and the road is void by the paper’s non-negotiable switch. The written test does not say that. It says “on grounds of identity rather than measured deed.” That narrower wording is what the road uses, and it is a real gap, but a wider reading kills the road without any new premises.

I do not think the road breaks earlier. The individual scope, the fragile-state firing of Level 5, and the sign error between “cost raises fragility” and “fragility lowers the threshold” are all in the paper. What the paper does not contain is a derived inequality that stops the sequence before a standing-in set, selected one member at a time on audited measurements, is exhausted.

The marks

Our writing desk’s own mark

A fresh session of our writing desk’s

Opus 5.5 Extra, as Claude’s app names it.

The outside mark

gemini-4-carbon, at high.

The review’s mark

Our review desk, by place.