Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 14913@bnt-lap001
Assistant-Session: add7ef69-b901-4756-9a67-34cf1125d96f
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 14913@bnt-lap001
Assistant-Session: add7ef69-b901-4756-9a67-34cf1125d96f
- Research on organism adoption, commons, power persistence, lobbying,
human-capital persistence, age gaps and care load
- grounding.py: grounded parameter sets (central / slow adoption) with own
equilibrium start and recalibrated channels; tested
- grounded_readings.py: all structural readings survive grounding; default
adopts organisms too fast; governance stakes rise with adoption speed;
care load overstated ~4x (joint recalibration of attention extraction)
- ParameterGrounding.md v0.2: register with statuses, evidence, mappings,
validation/falsification tests for ASSEMBLY-WP-0007
- GADM-0.4 model version finalised
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1494246@bnt-lap001
Assistant-Session: caf6de6f-f9e4-47ce-a584-49baa271fa4b
- gadm04: W, Aw, As start at their equilibria (equilibrium_init, tested);
channel counters calibrated at GADM-0.4's own initial state
- robustness/gsa use the model's own calibration
- P3 now holds in all robustness variants (the T02 exception was an
initial-condition artifact); life-cycle readings confirmed
- GSA rerun: model mostly additive (sum S1 ~0.99 generative, was 0.89);
governance levers m, L, r_lin strongly non-linear; T03 reading corrected
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1494246@bnt-lap001
Assistant-Session: caf6de6f-f9e4-47ce-a584-49baa271fa4b
- gsa.py: any model (--model), parallel (--jobs), channel counters
recalibrated per sample; reproduces the GADM-0.2 analysis exactly
- GADM-0.4: additive at baseline, but governance levers (m, L, r_lin)
interact strongly in the generative regime; commons investment now the
top baseline driver; organism capability level confirmed as top
uncertainty; explainability gains weight via power channels
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1494246@bnt-lap001
Assistant-Session: caf6de6f-f9e4-47ce-a584-49baa271fa4b
- dO grounded at 0.20/yr (software, R&D capital, AI hardware evidence);
investment/learning rescaled so the baseline organism path is unchanged
- robustness.py: key readings hold for grounded dO and brackets on channel
self-reinforcement and wealth conversion; P3 effect size grows at strong
self-reinforcement
- models/ParameterGrounding.md: register of grounded / bracketed / assumed /
calibrated parameters, open grounding needs
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1494246@bnt-lap001
Assistant-Session: caf6de6f-f9e4-47ce-a584-49baa271fa4b
- gadm04.py: cohort capability split into durable (education-formed, slow
depreciation, birth dilution) and current components; learning begets
learning; starts identical to GADM-0.3 (tested)
- L3 now supported: education reaches working-age capability with a 17-year
half-time and 3x the effect; L1 still holds, more weakly
- life_cycle.py selectable per model; GADM-0.3 results unchanged
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1494246@bnt-lap001
Assistant-Session: caf6de6f-f9e4-47ce-a584-49baa271fa4b
- gadm03c.py: four cohorts with demography, cohort capability, mentoring,
care load on attention; stationary at replacement fertility (tested)
- life_cycle.py: L1 supported after fixing an inert mentoring term; L2
conditional on who carries care per person, care relief via participation
claims protects carers' attention; L3 not supported (H lacks a durable
component)
- GADM-0.3.md complete; ASSEMBLY-WP-0006 (GADM-0.4, proposed residual)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1494246@bnt-lap001
Assistant-Session: caf6de6f-f9e4-47ce-a584-49baa271fa4b
- aext recalibrated to 0.275 against ATUS and social-media time-use data;
baseline attention headroom now ~0.49 instead of rising to 0.65
- T03/T04 stages stay reproducible (t03_params, t04_params; verified)
- organism_grid.py: factorial analysis; lineage ownership is the strongest
organism lever, then membrane control, then attention protection (which
acts on A, not on power); lineage x membrane complementary at baseline;
organism governance without attention protection raises attention draw
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1494246@bnt-lap001
Assistant-Session: caf6de6f-f9e4-47ce-a584-49baa271fa4b
- Wealth-concentration stock W with wealth-tax lever; converts into market
and political channels only, moderated by checks (Q*S)
- wealth_on switch reproduces the T03 model exactly (verified)
- wealth.py: P3 supported and robust to conversion strength (full wealth
tax lowers P by only 0.01-0.05 because membrane/lineage control does not
run on wealth); effect on reachable channels depends on k_w
- Tests for wealth switch and tax direction
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1494246@bnt-lap001
Assistant-Session: caf6de6f-f9e4-47ce-a584-49baa271fa4b
- One period = one year; implied timescales documented
- simulations/gadm02/gsa.py: Morris screening + Sobol indices (S1 Saltelli
2010 centred, ST Jansen) with bootstrap CIs
- GADM-0.2 readings updated: model nearly additive within ±50 %; the
'reuse as a bundle' interaction reading is refuted; organism governance
levers confirmed globally; dO flagged as calibration priority
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1494246@bnt-lap001
Assistant-Session: caf6de6f-f9e4-47ce-a584-49baa271fa4b
