It can help. It can be inherited. It can still disappear.
Working manuscript · v0.1
AN OPPORTUNITY THAT VANISHED
This world tests two inherited changes: a program that processes new food, and a setting that increases food intake. In EX-12, keeping or undoing them showed that higher intake helped when the program was present. Yet another birth replaced that organism before it reproduced. Its original history ended without offspring.
FOUR VERSIONS
A carries neither change. B carries the resource-processing capability. C carries high uptake without the capability. BC carries both. Each candidate was tested from its actual birth-time ecology and in standardized worlds, with ten paired futures per context and further mutation disabled.
ONE SUPPORTED OPPORTUNITY
One of four EX-12 candidates passed both contexts. B beat A in 9/10 futures, BC beat B in 10/10, and C-only did no better than A in 10/10 in each context. B was 45 ticks old and occupied 12.5% of slots at the C birth. Three independent qualifying discoveries were required; one did not meet that replication gate.
A DIFFERENT ORIGIN
EX-13 used fresh discovery worlds and followed the first clean candidates prospectively. One of five passed both birth-opportunity contexts. This time it survived, reproduced and transmitted both changes. Its intact descendants numbered 217 at +50 ticks, 148 at +200 and 230 at +500.
THE LATER TEST
The lowest-ID intact descendant at the registered +200 checkpoint was assayed in its actual ecology. The dependency effect was positive in six of ten paired futures. In the other four, every version produced zero offspring. The required eight wins were not reached. Full-gate retention stayed 0/1 supported opportunities; genetic persistence did not override that result.
A LINEAGE IS NOT A CAPABILITY
The selected event lineage went extinct at tick 1268. The world still contained 239 capable high-uptake organisms in other lineages, and 235 at the final tick. Those organisms were not counted as rescue of the selected origin.
EX-14 / THE CHANCE TO REPRODUCE
A separately registered test returned to that checkpoint with 100 paired futures. BC produced lineage births in 64/100. In the 64 futures where either B or BC reproduced, BC left more lineage births in 63. The other 36 futures remain in the unconditional denominator. The conditional mean gain was 99.45 lineage births; the median was 4.
SOURCE DISCREPANCY
The database conclusion and written report give B as reproducing in 41/100 and the remaining conditional case as a loss. The local ex14.jsonl gives B as 22/100, one reproductive tie and no loss. Both give BC as 64/100 and the 63/64 positive result. The figure uses the local raw file; both source accounts are preserved and their discrepancy remains unresolved.
WHAT THIS CHANGES
The later assay's weak result is consistent with a bottleneck in reproductive opportunity, rather than evidence that the conditional advantage reversed. This is one transmitted opportunity and one checkpoint ecology. The 100 futures are not 100 independent innovations, and EX-14 does not retroactively change EX-13's acceptance gate.
Can the whole sequence happen again?
EX-15 seeks at least three independently discovered opportunities through the full pipeline. As checked on 10 September 2026, it remains running with no reported outcome. A further dependent step, historical necessity and environmental mediation remain untested.
Lab notebook / Working draft
Conclusion
One family transmitted both changes before dying out. The later comparisons make reproductive opportunity a plausible bottleneck, rather than establishing that the conditional advantage reversed.