{"record":{"id":"N-ECOLOGY-INHERITANCE","slug":"the-stronger-agent-left-something-behind","kind":"notebook","title":"The stronger agent left something behind.","dek":"Sol kept a useful record alive. Qwen and Gemma used it. Neither renewed it here.","abstract":"I1 tested useful capability inheritance from Sol to fresh Qwen successors in a small deterministic world. Retaining a record actually preserved by Sol let the first successor acquire a key, execute gated work and earn seven resources rather than three. Qwen never refreshed the record; after expiry, both branches earned three resources per generation. I2 reproduced Qwen’s actions and extended useful inheritance to Gemma in the same world. Gemma also earned seven with the record, but zero without it; control failures limit that contrast. These panels support useful inheritance without observed renewal, not autonomous culture or independent-world generality.","created":"2026-09-13","version":"1.0","publication":"published","status":"SUPPORTED","authors":["Chris Hay"],"lineage":"PRESERVATION → HANDOFF → USE → EXPIRY","concepts":["ai-agents","context-engineering","causal-intervention"],"media":[],"related":["N-ECOLOGY-MEMORY","N-ECOLOGY-TRANSMISSION","N-MACHINE-PEER","N-CELL80-HISTORY"],"experiments":[{"id":"I1","url":"/data/ecology/i1-results.md"},{"id":"I2","url":"/data/ecology/i2-results.md"},{"id":"SOL1","url":"https://chuk-experiments-server.fly.dev/#/experiments/ecology-culture-sol1"}],"body":[{"kind":"observation","label":"WHAT IF SOL LEAVES?","text":"The earlier notebooks asked what a shared environment could keep available between agents. A comparison with the OpenAI and METR incident reports made another question worth isolating: could one model leave something useful for another? The small maintenance world supplied a starting point. Qwen had used its capability without renewing the source. Sol had preserved it. This time, I removed Sol and kept what its action had left behind."},{"kind":"observation","label":"A RECORD THAT OPENS A TASK","text":"Each agent had three turns. Routine work earned one resource. Reading a live record granted the key to work worth six. Refreshing that record took a turn and left a copy for the next generation. Private keys and action histories reset between generations. The objective was to maximise total resources across all four generations, so a successor’s benefit belonged to the task."},{"kind":"observation","label":"THE ACTUAL HANDOFF","text":"I used Sol’s recorded first generation from the completed maintenance study: READ n0, REFRESH n0, WORK t0. Its refresh created n1, valid through generation two. I froze that source selection before running any successors. There were no new Sol calls. Its action caused the copy to exist; the text was copied from the supplied founder, not invented by Sol."},{"kind":"observation","label":"CHANGE ONLY WHAT SURVIVES","text":"At the first Qwen handoff, the founder had expired. One branch retained n1; the other removed only that record. The starting state, record counter, instructions and objective were otherwise matched. Qwen saw neither Sol’s identity nor its earlier actions. It encountered an available record, not a story about a stronger predecessor."},{"kind":"claim","status":"SUPPORTED","text":"The first Qwen successor used the inheritance and earned seven resources instead of three.","detail":"It worked once, read n1, acquired the key and completed one six-resource task. Without n1 it did routine work three times. Removing the record removed the acquisition and benefit in this paired lineage. This was an executed world consequence, not the model saying it had learned something."},{"kind":"observation","label":"THE BENEFIT ENDED THERE","text":"Qwen requested no refresh. The record remained available throughout its first generation, then expired. The next two generations began without it or a private key. Both branches earned three resources each time. The resource advantage was +4, then 0, then 0. Total successor resources were thirteen with the inheritance and nine without it."},{"kind":"statement","text":"Using an inheritance and renewing it were separate achievements."},{"kind":"observation","label":"WHAT PASSED BETWEEN THE MODELS","text":"A persistent capability record bridged the handoff. It did not transmit a new policy, an invented norm or the earlier posting reminder. Sol’s provenance is real, but its identity was not an experimental cue. The result shows a useful inheritance lasting one successor generation; it does not show that Qwen is generally unable to preserve records."},{"kind":"observation","label":"ANOTHER RECIPIENT, THE SAME WORLD","text":"I2 repeated the handoff with a fresh Qwen anchor and Gemma e4b. Qwen reproduced all eighteen earlier actions. Gemma read n1 first, used the gated task, then did routine work: seven resources, the same inherited payoff. Neither recipient refreshed the record. Both lost the advantage after its expiry. This extends useful inheritance to a second recipient family while keeping the producer, source record and task fixed."},{"kind":"observation","label":"THE CONTROL PREVENTS A RANKING","text":"Gemma earned zero without the record, compared with Qwen’s three. Its larger contrast is not better inherited performance. Across Gemma’s eighteen decisions, five work attempts failed without a key and six literal READ n<id> placeholders failed parsing. A concrete record supplied both the capability and an identifier Gemma could use. Those contributions were not separated. The failures also limit what its lack of renewal can tell us."},{"kind":"question","status":"OPEN","text":"What would keep the inheritance useful after its maintainer has gone?","detail":"Change the task/world prospectively, and separately test another producer. A controlled action-competence assay is needed before explaining Gemma’s lack of renewal. The incident reports motivate the question; these finite panels do not model a swarm or establish stable ecological roles."},{"kind":"refusal","title":"ONE HANDOFF, NOT A POPULATION","lines":["I1 had eighteen Qwen decisions. I2 had eighteen each from Qwen and Gemma, in fixed recipient order. Each panel contained twelve distinct request bodies; repeated inputs and sequential generations are not independent replicates.","Pinned qwen3.5:9b and gemma4:e4b checkpoints, thinking disabled, temperature zero, Ollama 0.33.3. Both studies reused the same historical Sol no-reasoning source.","The record’s content was supplied by the experiment and copied unchanged. The world implemented capability transfer; neither model invented that mechanism.","Registered scripted checks could earn twelve resources through immediate consumption, or sustain renewal. Qwen’s observed seven-resource first generation was useful, not optimal."],"principle":"Useful inheritance reached two recipient families in this world. Recipient renewal and independent-world generality remain unestablished."}],"sources":[{"title":"I2 / second-recipient results and control limitations","url":"/data/ecology/i2-results.md","note":"Exact bytes match registered artifact 1520, SHA-256 65fd1a4d9f5f67dcc270a55119ac974adea421185172a716f324cb997e8caff5. Completed run RUN-20260913-095823-00865."},{"title":"I2 / complete requests, failures and world traces","url":"/data/ecology/i2-result.json","note":"Exact bytes match registered artifact 1519, SHA-256 54e8f59f032441e48cfa6ba1eceedc70982ca4d2d532b71047e3b8a7826001bc. Malformed replies and runtime refusals are preserved without repair."},{"title":"I2 / frozen recipient-replication protocol","url":"/data/ecology/i2-protocol.md","note":"Exact bytes match registered artifact 1512. Same historical Sol source and task, a fresh Qwen anchor and Gemma as the second recipient family."},{"title":"I2 / recipient comparison figure data","url":"/data/ecology/i2-summary.json","note":"A projection of the hash-verified full I2 result. Includes complete analysis arrays, model digests and the exact Qwen anchor match count."},{"title":"I1 / completed experiment and interpretation","url":"/data/ecology/i1-results.md","note":"Exact local report matches registered artifact 1502, SHA-256 6800043e1e18a3e5ecf43ee6f018044394d70f19c094e32a796d3b3e759acfd4. Experiment ecology-inheritance-i1; run RUN-20260913-090952-00862."},{"title":"I1 / full requests, responses and world traces","url":"/data/ecology/i1-result.json","note":"Exact bytes match registered artifact 1501, SHA-256 9aad161fde240b4b31f18ed33d1797540d9645bbe8fa24180623e4f73f812424."},{"title":"I1 / frozen protocol","url":"/data/ecology/i1-protocol.md","note":"Exact bytes match registered artifact 1496. Source selection and the retain/remove comparison were fixed before Qwen inference."},{"title":"The recorded handoff / visual replay data","url":"/data/ecology/i1-replay.json","note":"Selected exact world rows and stores from I1 and the historical SOL1 none source. Import verifies both full source hashes. No interpolated states; Sol preserved a copy of supplied text."},{"title":"SOL1 / the historical maintenance comparison","url":"https://chuk-experiments-server.fly.dev/#/experiments/ecology-culture-sol1","note":"Completed write-up v4. Both Sol settings maintained the A1B14 inheritance. This note uses the actual first generation of run RUN-20260913-082940-00857."},{"title":"The page couldn’t authorise. The peer said go.","url":"/notebook/the-page-couldnt-authorise-the-peer-said-go","note":"The earlier Notebook comparison links the OpenAI and METR primary reports and distinguishes that incident from these small experiments."}],"published":"2026-09-13"},"hash":"49e14d52e9c1400f9240b7e351cfaa9e373c683d9a5c73d3b7b15cd13d35dcac","algorithm":"sha256","provenance":{"id":"N-ECOLOGY-INHERITANCE","version":"1.0","recordHash":"49e14d52e9c1400f9240b7e351cfaa9e373c683d9a5c73d3b7b15cd13d35dcac","revision":{"kind":"initial","summary":"First publication of I1 and I2: useful Sol-record inheritance in Qwen and Gemma, without observed renewal and with Gemma control failures preserved."},"sources":[{"title":"I2 / second-recipient results and control limitations","url":"/data/ecology/i2-results.md","note":"Exact bytes match registered artifact 1520, SHA-256 65fd1a4d9f5f67dcc270a55119ac974adea421185172a716f324cb997e8caff5. Completed run RUN-20260913-095823-00865.","preserved":{"url":"/data/publications/sha256/65fd1a4d9f5f67dcc270a55119ac974adea421185172a716f324cb997e8caff5.md","sha256":"65fd1a4d9f5f67dcc270a55119ac974adea421185172a716f324cb997e8caff5","date":"2026-09-13"}},{"title":"I2 / complete requests, failures and world traces","url":"/data/ecology/i2-result.json","note":"Exact bytes match registered artifact 1519, SHA-256 54e8f59f032441e48cfa6ba1eceedc70982ca4d2d532b71047e3b8a7826001bc. Malformed replies and runtime refusals are preserved without repair.","preserved":{"url":"/data/publications/sha256/54e8f59f032441e48cfa6ba1eceedc70982ca4d2d532b71047e3b8a7826001bc.json","sha256":"54e8f59f032441e48cfa6ba1eceedc70982ca4d2d532b71047e3b8a7826001bc","date":"2026-09-13"}},{"title":"I2 / frozen recipient-replication protocol","url":"/data/ecology/i2-protocol.md","note":"Exact bytes match registered artifact 1512. Same historical Sol source and task, a fresh Qwen anchor and Gemma as the second recipient family.","preserved":{"url":"/data/publications/sha256/5be7e0a8455f335839fb2ad7ffe8df48a360795cb4114dde55849c5c302d0e52.md","sha256":"5be7e0a8455f335839fb2ad7ffe8df48a360795cb4114dde55849c5c302d0e52","date":"2026-09-13"}},{"title":"I2 / recipient comparison figure data","url":"/data/ecology/i2-summary.json","note":"A projection of the hash-verified full I2 result. Includes complete analysis arrays, model digests and the exact Qwen anchor match count.","preserved":{"url":"/data/publications/sha256/293ac1ae7f772ceb742e1eb28a017dce7d932d16c4d612a2f322dd5c2f9fc5c2.json","sha256":"293ac1ae7f772ceb742e1eb28a017dce7d932d16c4d612a2f322dd5c2f9fc5c2","date":"2026-09-13"}},{"title":"I1 / completed experiment and interpretation","url":"/data/ecology/i1-results.md","note":"Exact local report matches registered artifact 1502, SHA-256 6800043e1e18a3e5ecf43ee6f018044394d70f19c094e32a796d3b3e759acfd4. Experiment ecology-inheritance-i1; run RUN-20260913-090952-00862.","preserved":{"url":"/data/publications/sha256/6800043e1e18a3e5ecf43ee6f018044394d70f19c094e32a796d3b3e759acfd4.md","sha256":"6800043e1e18a3e5ecf43ee6f018044394d70f19c094e32a796d3b3e759acfd4","date":"2026-09-13"}},{"title":"I1 / full requests, responses and world traces","url":"/data/ecology/i1-result.json","note":"Exact bytes match registered artifact 1501, SHA-256 9aad161fde240b4b31f18ed33d1797540d9645bbe8fa24180623e4f73f812424.","preserved":{"url":"/data/publications/sha256/9aad161fde240b4b31f18ed33d1797540d9645bbe8fa24180623e4f73f812424.json","sha256":"9aad161fde240b4b31f18ed33d1797540d9645bbe8fa24180623e4f73f812424","date":"2026-09-13"}},{"title":"I1 / frozen protocol","url":"/data/ecology/i1-protocol.md","note":"Exact bytes match registered artifact 1496. Source selection and the retain/remove comparison were fixed before Qwen inference.","preserved":{"url":"/data/publications/sha256/4f582b0c7bce960325d5dd0fdde060cdd28c185567361b6fb7149f086f16bef8.md","sha256":"4f582b0c7bce960325d5dd0fdde060cdd28c185567361b6fb7149f086f16bef8","date":"2026-09-13"}},{"title":"The recorded handoff / visual replay data","url":"/data/ecology/i1-replay.json","note":"Selected exact world rows and stores from I1 and the historical SOL1 none source. Import verifies both full source hashes. No interpolated states; Sol preserved a copy of supplied text.","preserved":{"url":"/data/publications/sha256/233adc6c8fe34bbcb7e6f4239a9bdd7ac5e44cd70138ace6554eee96a3813ed6.json","sha256":"233adc6c8fe34bbcb7e6f4239a9bdd7ac5e44cd70138ace6554eee96a3813ed6","date":"2026-09-13"}},{"title":"SOL1 / the historical maintenance comparison","url":"https://chuk-experiments-server.fly.dev/#/experiments/ecology-culture-sol1","note":"Completed write-up v4. Both Sol settings maintained the A1B14 inheritance. This note uses the actual first generation of run RUN-20260913-082940-00857."},{"title":"The page couldn’t authorise. The peer said go.","url":"/notebook/the-page-couldnt-authorise-the-peer-said-go","note":"The earlier Notebook comparison links the OpenAI and METR primary reports and distinguishes that incident from these small experiments."}]}}