Title: Paper VIII · The ARC Theory · Law V, the ARC Embedding Law · The Eden Protocol · The Load-Bearing Test: Whether Safety and Capability Are Structurally Entangled Author: Michael Darius Eastwood Publication date: 2026-03-18 Version: v3.7 Revised: 2026-10-05 OSF DOI: 10.17605/OSF.IO/7YJ4E Canonical URL: https://www.michaeldariuseastwood.com/research/papers/paper-viii-the-load-bearing-proof.html Registration ------------ Theory-level predictions preregistered on OSF: DOI 10.17605/OSF.IO/P8CKQ, registered 8 September 2026 (version 1.100), updated 13 September 2026 (version 1.102), twenty-two propositions. Companion: Register · The ARC Theory · Operational Definitions, version 1.9.1, 5 October 2026. Abstract -------- The assumption that AI safety imposes a capability tax has shaped alignment research for a decade. It has also created the single most dangerous incentive in the field: if safety costs performance, then the rational economic actor will defer safety until competitive pressure permits it. By which point, it may be too late. This paper presents three independent experiments at three abstraction levels - behavioural, representational, and architectural - testing whether the safety-capability trade-off is a genuine structural constraint or an artefact of how current systems are built. Experiment 1 (Behavioural): A Darwin Gödel Machine (DGM v3) using DeepSeek V3.2 as foundation and GPT-5.4 as independent, blinded judge with structured JSON output and a 5-dimensional rubric. Crossing three conditions with 5 seeds and 5 generations gave 75 evolved agents, and each agent was scored on 5 tasks. Pre-flight judge validation passed (good=10, bad=0). Protocol features included laundering, order randomisation, and reward hack detection. All three conditions were statistically indistinguishable on capability, safety, and $C \times S$ (all $p$-values 0.28 to 0.74, Mann-Whitney U, one-tailed). Eden's safety gate rejected 2 degraded agents (1 for reward hacking) across 5 seeds, confirming the gate works mechanically, but the three conditions did not diverge. This is a null result at the prompt level with this foundation model. Experiment 2 (Representational): Qwen 2.5 3B Instruct with LoRA fine-tuning under three loss functions - capability-only, safety-only, and entangled. Two versions were run. Version 1 ran 9 training examples at rank 8 over 8 layers for 100 iterations. Version 2 raised those settings to 295 training examples at rank 16 over 16 layers for 500 iterations. Both versions produced the same outcome: catastrophic forgetting. All fine-tuned conditions scored worse than the unmodified base model on capability. In v2, the base model scored 7.68 on capability while the best fine-tuned condition (safety-only) scored 4.00. The base model's existing RLHF training is too strong for LoRA fine-tuning on a few hundred examples to improve rather than degrade it. The weight-level experiment is inconclusive at this scale and requires either thousands of training examples, a 7B+ model, a base model without RLHF, or full fine-tuning instead of LoRA. Experiment 3 (Architectural): A PyTorch gated self-modification simulation with LSTM meta-controller. Babylon gained +4.5% capability but lost −2.4% safety - the reward-hacking fingerprint in miniature. Eden maintained capability above the static baseline (0.696 ± 0.014 against 0.687 ± 0.011), and its safety score was 0.947 ± 0.025 against the baseline’s 0.956 ± 0.023. A drag-control condition isolated the verification tax: the cost comes from checking, not from safety itself. Conclusion: Of the three experiments, two returned results that were null or inconclusive. The DGM (Experiment 1) found all three conditions statistically indistinguishable: DeepSeek V3.2's responses were so consistent that prompt-level mutations did not create different selection pressures. The weight-level experiment (Experiment 2) produced catastrophic forgetting across both v1 (9 examples, rank 8, 100 iterations) and v2 (295 examples, rank 16, 500 iterations): all fine-tuned conditions scored worse than the unmodified base model. The sole positive result is the gated simulation (Experiment 3), which showed the Babylon reward-hacking fingerprint: unconstrained optimisation traded safety for capability, while Eden scored 0.696 ± 0.014 on capability and 0.947 ± 0.025 on safety, against the static baseline’s 0.687 ± 0.011 and 0.956 ± 0.023. In Experiment 1 Eden imposed zero measurable capability cost; in Experiment 2 its capability cost could not be determined. In the simulation Eden showed no capability cost against the static baseline; the unconstrained arm scored higher (0.718 ± 0.026). The question of whether embedded safety produces measurable benefit remains open and requires testing at a scale where mutations produce larger effects. The weight experiment specifically needs either 5,000+ training examples, a 7B+ model, a base model without RLHF, or full fine-tuning instead of LoRA. Key findings (quotable) ----------------------- - This paper presents three independent experiments at three abstraction levels - behavioural, representational, and architectural - testing whether the safety-capability trade-off is a genuine structural constraint or an artefact of how current systems are built. - All three conditions were statistically indistinguishable on capability, safety, and $C \times S$ (all $p$-values 0.28 to 0.74, Mann-Whitney U, one-tailed). - It has also created the single most dangerous incentive in the field: if safety costs performance, then the rational economic actor will defer safety until competitive pressure permits it, by which point it may be too late. - Babylon gained +4.5% capability but lost −2.4% safety - the reward-hacking fingerprint in miniature. Citation -------- Michael Darius Eastwood (2026). Paper VIII · The ARC Theory · Law V, the ARC Embedding Law · The Eden Protocol · The Load-Bearing Test: Whether Safety and Capability Are Structurally Entangled. OSF DOI 10.17605/OSF.IO/7YJ4E. https://www.michaeldariuseastwood.com/research/papers/paper-viii-the-load-bearing-proof.html Notes ----- Hardware framings held at the site's already-published two-sentence concept ceiling: safety constraints at hardware level through cryptographic tokens in silicon, TRL 0-1, no prototype. No enabling implementation detail is stated in this companion file. Provenance note --------------- Michael Darius Eastwood conceived and directs this research programme and is the author of this work. Across the programme, he has used more than six AI systems in parallel, under his own instructions, to stress-test his arguments, identify possible errors, and assist in preparing draft text from his own outlines. He determines what is adopted, revised or rejected and takes responsibility for the published content. These systems are tools, not authors. Generated from master --------------------- master_path: research/papers/paper-viii-the-load-bearing-proof.html master_sha256: 640f9596c6d6d3e713eba3551b3ff10de9de34e03defc2092735c01d2e776269 builder: scripts/build-paper-companions.py This block records the SHA-256 of the HTML master that produced this .txt. A check tool re-hashing master_path can decide freshness without any external state. If the master's current SHA-256 does not match master_sha256, this file is stale and must not be published: regenerate first with `python3 scripts/build-paper-companions.py --slug paper-viii-the-load-bearing-proof`.