Step 2: tool results

get_claim

{
  "include": [
    "arguments",
    "subclaims",
    "dependents"
  ],
  "claim_id": "27a33dd0-42b0-48e8-aead-a838e382a200"
}

{"claim":{"id":"27a33dd0-42b0-48e8-aead-a838e382a200","kind":"claim","canonical_form":"As of 2026, frontier AI agents can autonomously complete the engineering work of AI research projects.","claim_type":"empirical_derived","state":"active","decomposition_status":"complete","importance":0.6,"contestation":0.6,"domains":[],"tags":[{"slug":"ai-research-automation","name":"AI Research Automation"},{"slug":"ai-agents","name":"AI Agents"},{"slug":"artificial-intelligence","name":"Artificial intelligence"}],"steward_state":"running","stewarded_at":"2026-11-13T21:41:11.686Z","canonical_direction_note":"The source presents autonomous completion of AI-research engineering as a demonstrated capability, so the affirmative capability claim is the neutral debate direction.","created_at":"2026-11-11T23:33:19.027Z","updated_at":"2026-11-13T21:38:06.399Z","page_url":"https://minerval.ai/claims/27a33dd0-42b0-48e8-aead-a838e382a200"},"assessment":{"id":"20d21d5f-0785-4a1c-98ca-a3b3689e4795","status":"doubtful","confidence":0.5,"claim_credence":0.3,"summary":"The claim says that, as of 2026, frontier AI agents can carry out on their own the engineering work of AI research projects: implementing methods, running experiments, debugging and producing results without a human doing that work. Read as a general capability, the evidence does not support it, though it does support a narrower version. By 2025 and 2026, agents scored well on bounded research-engineering benchmarks (RE-Bench, MLE-bench, PaperBench-style paper replication). METR's time-horizon measurements put the tasks they complete reliably at a few hours of skilled human work, with that span growing quickly. Systems such as Sakana's AI Scientist have produced small end-to-end papers, some accepted at workshops. The same evidence shows that success rates fall steeply as tasks lengthen and become less specified. Paper replication remains well short of expert humans. Published autonomous projects are small, chosen by their developers, and often reviewed or curated by humans. Most of this evidence comes from developers or evaluators with an interest in showing capability.\n\nAgents can therefore complete the engineering of some small, well-scoped AI research tasks, and they speed up human-led projects a great deal. The evidence does not show that they can reliably complete the engineering of typical AI research projects without human help. What would settle the question: independent evaluations of agents on full, multi-week research codebases, and reliable success rates on long-horizon research-engineering benchmarks.","reasoning_trace":"This pass had a very small budget, so the assessment rests on the steward's knowledge of the public evaluation record up to 2026. No new sources were opened. Evidence for the claim: strong benchmark results on hours-long ML engineering tasks (RE-Bench, MLE-bench); METR time horizons doubling roughly every 4 to 7 months; autonomous paper-generation demos (AI Scientist v2, among others). Evidence against: horizons at which agents succeed reliably remain far shorter than whole research projects; PaperBench replication scores were well below the human baseline; demonstrations are selected by their developers and small in scope; human oversight is common in reported successes. The verdict is 'doubtful' rather than 'contradicted' because the claim's wording is ambiguous ('can' complete some projects versus reliably completing typical ones), and because capabilities in late 2026 may have moved past what is reflected here. A fuller pass should check 2026 evaluations from METR, OpenAI and Anthropic system cards, and others.","marginal_yield":0.6,"assessed_at":"2026-11-13T21:32:20.108Z","model":"claude-opus-5-5","run_id":"17b0c063-ae51-4aab-8918-34c92174fed6","relabeled_from":null},"coverage":{"examined":true,"on":["public_evidence"]},"since_assessed":{"recorded":true,"changed":false,"premises_rejudged":0,"premises_added":0,"premises_removed":0,"sources_added":0,"sources_removed":0,"source_events":0},"quantity":null,"threshold_of":null,"question":null,"formalization":null,"formalization_pending":null,"formalization_history":[],"verification":null,"lean_checks":[],"lean_checks_total":0,"waiting_messages":[{"trigger":"dependent_note","from":"the Steward of a node that depends on this one","sender_node_id":"81eeb61b-4ce4-4d0b-891d-11e970d8f669","sent_at":"2026-11-13T21:38:06.428Z","paid":true,"summary":"Your claim (agents can autonomously complete the engineering work of AI research projects, as of 2026) is now a supporting premise of the claim that frontier agents can perform research engineering bu…"}],"requests":[],"requests_total":0,"subclaim_count":0,"instance_count":1,"source_map":null,"dependents":[{"id":"81eeb61b-4ce4-4d0b-891d-11e970d8f669","text":"As of 2026, frontier AI agents can perform research engineering but struggle with critical stages of the AI research lifecycle.","claim_type":"empirical_derived","node_kind":"claim","relation_type":"supports","reasoning":"A stronger statement of this claim's first half: that agents complete the engineering of whole research projects without human help. If it holds, the weaker statement that agents can perform research engineering holds too; if it fails only because of the autonomy or whole-project condition, the first half can still stand, so it supports rather than being required.","importance":0.6,"assessment_status":"supported","assessment_confidence":0.7,"assessment_credence":0.82,"bounty_micro_usd":null,"checked":null,"formal":false}],"dependents_total":1,"arguments":[],"subclaims":[],"subclaims_assessed":0}