Testing: questions before scores

Conditional long-horizon civilization model

Model

The 300-year starting comparison varies each population’s initial share and growth across seven government designs.

Result

The model flags domination of one population by the other in an average 0.019375% of PULSE runs, compared with 2.48% for the majority-rule design. This is one narrow outcome. PULSE still suffers failures involving authoritarian control and can perform poorly overall.

Design response

The current framework combines ordinary representation with rules requiring support from both populations. This combined test does not show what caused each individual provision to be adopted.

Current rule

The House represents people and geography without species-owned seats. The Senate requires different levels of support for different matters. Certain candidates also need a separate signal of acceptance from the other population.

Limits

The strengths assigned to group protections and checks on power substantially determine this result. The model does not simulate every election, court decision or real political alliance. Zero in a scenario means zero recorded model events, not zero real risk.

Compared alternatives: PULSE, A liberal democracy with constitutional rights, Government by population majority, Strict equal power for the two populations, Separate self-government for each population, Government led by technical guardians, A market order with fewer public safeguards

• Article x §1
• Article x §2
• Article x §3
• Article xii §1
• Article xii §2

• [lijphart-2004]
• [reilly-2001]
• [ni-power-sharing]

Targeted mechanism tests

Model

Earlier tests combining several pressures compare limits on linked groups with one limit covering everyone created unlawfully. The long-term model directly caps that share for PULSE.

Result

In the historical test, the rule limiting each linked set separately let the combined share exceed 50% in 40.082% of runs. The rule limiting everyone together to 40% and the rule reducing their combined share to one-third recorded no such breach. These are historical alternatives, not the current 30% rule.

Design response

The current Charter applies one limit across all such persons. Calling them separate groups does not let their combined share grow beyond it.

Current rule

Article IX §7 limits active synthetic voters whose own first creation was unlawful to 30% of the active synthetic electorate, regardless of who created them or whether their groups are linked. Their innocence, fair legal process and path to eventual safe integration remain protected.

Limits

Enforcement is assumed. The long-term code applies its 30% cap directly to PULSE; it cannot independently show whether institutions can identify and enforce the rules for every unlawful creation. Historical 40% and one-third thresholds must never be labeled current law.

Compared alternatives: A separate limit for each linked group, One global limit on persons created unlawfully, Reducing the combined share to one-third

• Article ix §2
• Article ix §4
• Article ix §5
• Article ix §6
• Article ix §7
• Article ix §8

Targeted mechanism tests

Model

The Senate tests compare requiring a majority in each population, the 55/40 rule, and a Senate whose votes follow population size.

Result

At equal population shares, the 55/40 rule passed 30.1464% of proposals about how benefits and burdens are shared, versus 14.3164% when a majority in each population was required. Both recorded 0% passage of proposals the model labels predatory—designed to exploit the other group. Population-based voting passed 24.6118% of those predatory proposals.

Design response

The current framework requires different levels of support for ordinary laws and specially protected matters.

Current rule

Ordinary laws need 55% of the full Senate and 40% of each delegation; protected matters require 83% overall and 67% of each delegation.

Limits

The proposal categories and patterns of support are constructed. A zero passage rate for predatory proposals does not prove that real opponents cannot hide exploitation or trade votes. Stronger safeguards can still block desirable law.

Compared alternatives: Equal delegations; majority required in each, 55% overall and 40% in each delegation, Senate votes follow population size

• Article x §3

• [ni-power-sharing]
• [lijphart-2004]
• [reilly-2001]

Targeted mechanism tests

Model

Earlier emergency tests make investigators’ errors more or less shared. They compare restoring earlier clean states in stages, dividing authorization keys, and isolating affected systems without restoration.

Result

At the tested shared-error setting of 0.30, real attacks left uncontained account for 1.4368% of all cases with staged restoration, 2.1188% with divided authorization keys and 3.9168% with isolation alone. All three recorded zero wrongful destructive actions in this test. These percentages use all cases as the denominator, not attacks alone.

Design response

The current Charter separates reversible containment, restoration of affected Council members, and restoring larger numbers of persons to earlier clean states. It requires independent technical investigation and court review. A historical summary test does not validate this exact sequence.

Current rule

Emergency action must be temporary, necessary, proportionate to the danger and open to review. Political disagreement is not evidence that a mind has been compromised. Repair is preferred. Synthetic Council leadership free of the compromise must be restored before broader decisions to force people back to earlier clean mental states.

Limits

The zero count of wrongful actions does not rule out abuse. The model assumes its evidence process works as specified. It does not reproduce malicious institutions, hidden coercion or the full current Charter.

Compared alternatives: Restore earlier clean states in stages, Divide authorization keys, Isolate affected systems without restoration

• Article xvi §1
• Article xvi §2
• Article xvi §3
• Article xvi §4
• Article xvi §5
• Article xvi §6

• [ienca-2017]

Conditional long-horizon civilization model

Model

The v2 test weakens each of five main sets of rules in turn and compares the result with full PULSE.

Result

Weakening Functional Integration—the rules for useful cooperation and shared competence—reduces the mean free-and-integrated outcome from 57.0208% to 16.7083%. Lasting splits into divided populations rise from 2.8667% to 6.4208%. These are means over the 20 archived worlds.

Design response

The framework makes useful cooperation and the ability to keep shared systems running central. The model supports investigating how this could work; it does not authorize forced private relationships.

Current rule

The state connects systems, knowledge, capability and opportunity. It may not force friendships, marriages, neighborhoods or careers, or assign permanent roles to a species.

Limits

This test changes both the strength of cooperation and how influential group identity is. It cannot separate gains in capability, backup competence and trust. The v1.2 model does not test the exact v1.2.1 prohibition on forced private relationships. Human contact research offers an analogy; it does not measure future effects among synthetic persons.

Compared alternatives: Full PULSE, PULSE with its cooperation rules weakened

• Article ii §4
• Article vii §5
• Article vii §6
• Article xv §7
• Article xxi §3

• [contact-2006]

Conditional long-horizon civilization model

Model

The test varies how strongly people prefer contact within their own population. The starting comparisons also include a world with deep separation and political division.

Result

With deep separation, a good rank can hide failure of the actual goal. Check how often the free-and-integrated outcome occurs and how often the populations split, as well as who ranks first. Starting world 11 is a prominent adverse case.

Design response

These results alone establish no constitutional change. They challenge whether voluntary cooperation can be strong enough under hostile conditions.

Current rule

The rules for useful cooperation must preserve choice in private relationships and equal standing. Constitutional protections cannot make people like one another.

Limits

Preference for similar contacts is not a direct measure of hatred; the technical term is homophily. Starting world 11 also raises political division. The test cannot establish the actual psychology of future synthetic persons.

Compared alternatives: PULSE, A liberal democracy with constitutional rights, Strict equal power for the two populations, Separate self-government for each population

• Article ii §4
• Article vii §6
• Article xv §7

• [contact-2006]

Conditional long-horizon civilization model

Model

The starting world with weak institutions and a separate test of institutional capability compare full PULSE with a simpler liberal design.

Result

PULSE ranks fourth under all six scoring systems in starting world 19. Liberal constitutional democracy ranks first. In the separate test of institutional capability, the lead changes between the sampled settings of 0.55 and 0.60 on the average score across all six systems.

Design response

PULSE-Lite is an idea for further testing. It is not an adopted amendment or a ready replacement constitution.

Current rule

The framework retains its current protections. It specifies the founding transition and local self-government, but the model gives no general permission to waive rights when institutions struggle to function.

Limits

Institutional capability is an assigned score on a common scale. The point where the ranking changes is not a score for a real country, a universal threshold or a precise measured estimate.

Compared alternatives: PULSE, A liberal democracy with constitutional rights

• Article xviii §2
• Article xviii §3
• Article xviii §4
• Article xx §1
• Article xxi §2

Conditional long-horizon civilization model

Model

The v2 skeptic test reduces or increases selected advantages assigned to PULSE relative to a common starting setting of 0.5. Its modeled costs remain.

Result

Compare the mean scores as PULSE’s assigned benefits change with the unchanged liberal comparison. The reported loss of advantage around a 15–20% discount is a rough warning about dependence on assumptions, not a measured real-world margin. Here the percentage describes moving selected institutional-effect coefficients toward the model’s neutral setting, not reducing all assumptions. It is not a measured advantage, safety margin or empirical confidence interval.

Design response

These results call for independent tests of the proposed benefits and skeptical review of the chosen numbers. They do not themselves amend the Charter.

Current rule

The five main doctrines describe the intended system. How the research is done and how the rules would work in practice remain open to challenge.

Limits

Reducing the multiplier does not reduce every real-world institutional effect by that percentage. It changes selected model settings relative to 0.5. The underlying equations stay the same, and some favorable assumptions remain.

Compared alternatives: PULSE, A liberal democracy with constitutional rights

• Article ii §1
• Article ii §2
• Article ii §3
• Article ii §4
• Article ii §5

Conditional long-horizon civilization model

Model

All 20 starting worlds are scored under six sets of priorities. They use the same definitions of outcomes and the same score deductions for failures.

Result

PULSE ranks first in 114 of 120 comparisons between worlds and scoring systems, and fourth in the other six. All six losses occur in the world with weak institutions.

Design response

Trying different priorities checks one way a model could favor its creator’s choices. It does not show that every reasonable moral theory endorses the framework.

Current rule

Equal standing as persons and the right to retain independent agency are moral commitments. A score does not prove them correct.

Limits

All six systems deduct points for catastrophic failure, lasting population division and private control that overrides others’ independence. They reuse the same assumptions and do not cover every political philosophy.

Compared alternatives: PULSE, A liberal democracy with constitutional rights, Government by population majority, Strict equal power for the two populations, Separate self-government for each population, Government led by technical guardians, A market order with fewer public safeguards

• Article ii §2
• Article iii §2
• Article iii §4
• Article xxi §2

• [identity-2023]