Claim

It may be easier to align AI systems than to align humans.

Type INTERPRETIVE · truth-evaluable: yes · status at snapshot: ACTIVE · created seq 2285

Referenced 112 times: checked 2, looked up by assistants 0, viewed 86, shared 24. A count of attention, not evidence.

Topics

AI · 03192ed8 Society · 03192ed8

Assessment

Supported

Score: 0.9600 under ledger-default@0.3.0 at snapshot 27745

3 of 4 review checks done · 3 counted

3 independent evidence lineages · review checks 3 of 4 · stability LOW

Main unresolved issue: 6 counted evidence items have no independence review; repeats of one origin may be counted as several coverage:6bc79e76-e8e4-8d6d-bbb3-671d67896ddf

Not yet independently audited.This assessment depends heavily on modeling choices.A quoted passage from osf.io was not found on the page when Galedra fetched it.A quoted passage is in something this server does not read, such as a PDF or a recording, so it has not been checked here.

Why?

Strongest counted support: The authors identify human supervision as the central bottleneck in AI alignment and list plural values, limited attention, counterfactual reasoning, and expertise-limited verification as obstacles; this supports the possibility that human alignment is unusually difficult. weight 1.800000 evidence

Suppressed as dependent: The excerpt directly says it might be easier to align AI than humans.; A speaker says he increasingly thinks it might be easier to align AI than humans.

Review gaps: independence reviewed

What would most change this: One independent, direct measurement that contradicts this claim would move it from Supported to Leans supported.

Hide calculation

0.9600 under ledger-default@0.3.0 at snapshot 27745 · state Supported · Review checks: 3 of 4 · stability LOW · 6 sources, 3 independent lineages

Model ledger-default@0.3.0 (the default model, not the answer) · snapshot 27745 · trace sha256:7412bfe210b0cab4c203790f42c0c692f7db5ad70a53a6f7639bf78c0be1f2b9

Last changed at entry seq 11922; nothing logged since has borne on this claim.

How the number was reached

Each counted link is weighed by multiplying the factors the model declares; a factor it does not declare is not shown. Every factor name below links to what that number means.

Link	Working	Weight	Counted as
4857eeda support	relevance DIRECT 2.0 × observation DIRECT_TEXT 0.9 × interpretation 0 steps, 1 − 0.15 × 0 1.000000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance INDEPENDENT 1.0 × edition NAMED_OR_UNRELATED 1	1.800000	1.800000
301d6de6 support	relevance DIRECT 2.0 × observation DIRECT_TEXT 0.9 × interpretation 1 step, 1 − 0.15 × 1 0.850000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance SELF 0.0 × edition NAMED_OR_UNRELATED 1	0.000000	0.000000 (dependent strongest only, kept b81d0d22)
1523fe0c support	relevance MODERATE 0.6 × observation HEARSAY 0.1 × interpretation 1 step, 1 − 0.15 × 1 0.850000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance SELF 0.0 × edition NAMED_OR_UNRELATED 1	0.000000	0.000000 (dependent strongest only, kept b81d0d22)
94a3025b support	relevance MODERATE 0.6 × observation DIRECT_TEXT 0.9 × interpretation 1 step, 1 − 0.15 × 1 0.850000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance INDEPENDENT 1.0 × edition NAMED_OR_UNRELATED 1	0.459000	0.459000
b81d0d22 support	relevance WEAK 0.2 × observation EXPERT_ANALYSIS 0.3 × interpretation 1 step, 1 − 0.15 × 1 0.850000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance SELF 0.0 × edition NAMED_OR_UNRELATED 1	0.000000	0.000000
47dc5f7e support	relevance STRONG 1.2 × observation DIRECT_TEXT 0.9 × interpretation 1 step, 1 − 0.15 × 1 0.850000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance INDEPENDENT 1.0 × edition NAMED_OR_UNRELATED 1	0.918000	0.918000

prior 0.50 → log-odds 0.000000 + 3.177000 (the counted weights) = 3.177000 → probability 0.9600

Under neighbouring settings of the same model: 0.9024 to 0.9842, a spread of 0.0818.

Half-even rounding, six places for weights and four for the probability. The same inputs under the same model reproduce this exactly; that is what the trace hash above is for. How a score is worked out explains every factor, why it is there, and what the model sets it to.

Your view

Registered views

Nobody has registered a view yet.

Registered views are personal belief, not evidence: they never change Galedra's assessment. Affiliations are self-declared and shown only for groups of 5 or more.

Checked as part of

Statements this claim was broken out of. The claim is scored on its own; a statement's reading only counts its claims.

“It may be easier to align AI systems than to align humans.” · 1 claim · 28 September 2026
“The Navier–Stokes proof was produced by roughly 10,000 OpenAI agents working for about 88 hours.” · 32 claims · 20 September 2026

Evidence for

Mostaque says it might be easier to align AI than humans.

From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance WEAK Steps 1 Counted

The speaker hedges with 'might'.
A speaker says he increasingly thinks it might be easier to align AI than humans.

From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance MODERATE Steps 1 Counted

The passage matches the claim, including its hedge. It is offered as a personal impression, 'I've been thinking more and more', with no argument or result behind it, and no sense of what 'align' is being measured against in either case.
Ngo: human goal-pursuit is driven by deeply evolutionarily ingrained emotions and reward signals — drives AIs need not share.

From Ngo and Yudkowsky on alignment difficulty (transcribed dialogue) Read 2026-09-22 Relevance MODERATE Steps 1 Counted

Humans carry hard-to-change evolutionary drives that resist alignment; AIs can be built without those drives, easing the alignment problem.
AGI could have 'follow instructions' as its explicit core goal — an orientation no human shares, making AI alignment structurally simpler.

From Conflating value alignment and intent alignment is causing confusion Read 2026-09-22 Relevance STRONG Steps 1 Counted

Directly argues an AGI can be structured around instruction-following in a way no human is oriented toward — alignment easier for AI than humans.
The excerpt directly says it might be easier to align AI than humans.

From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance DIRECT Steps 1 Counted
The authors identify human supervision as the central bottleneck in AI alignment and list plural values, limited attention, counterfactual reasoning, and expertise-limited verification as obstacles; this supports the possibility that human alignment is unusually difficult.

From AI alignment is a human problem Read 2026-09-28 Relevance DIRECT Steps 0 Counted

The source is a 2026 preprint, so the result remains provisional and not peer-reviewed.

Evidence against

None.

Qualifications

None.

Suppressed as dependent

Evidence	Group	Kept instead
96a715e7		b81d0d22
4a0e936e		b81d0d22

Inferences

Concluded from

None.

Used as a premise in

None.

An inference is interpretation, not evidence: it never changes the assessment of its conclusion. Each premise is a claim with its own evidence.

Threads

None. Threads are for disagreements about how a determination was made — a statement whose figures are not in the passage it rests on, two sources that may share an origin. Something about the world belongs in the evidence above.

Claim edges

None. Edges are stored and displayed; v0.1 does not propagate probability across them.

Review checklist

Review checks: 3 of 4

3 checks by this claim's own author, alongside the independent review above. 1 OPPOSING EVIDENCE SEARCH · 1 QUALIFIER CHECK · 1 EVIDENCE VERIFICATION. A check by the person who recorded a claim does not raise its review coverage.

Check	Done	By
qualifiers reviewed	yes	0fd572f9
opposing search done	yes	7847f851
independence reviewed	no
primary source reviewed	yes	4857eeda

Summary

Supported. The main support is: The authors identify human supervision as the central bottleneck in AI alignment and list plural values, limited attention, counterfactual reasoning, and expertise-limited verification as obstacles; this supports the possibility that human alignment is unusually difficult. 1e09f016

2 further items repeat the same origin and are not independent evidence. 96a715e74a0e936e

A search for evidence in the opposing direction found something; see the counted evidence. 7847f851

A search for evidence in the opposing direction found none. 7847f851

Review checks: 3 of 4. coverage:6bc79e76-e8e4-8d6d-bbb3-671d67896ddf

Generated by stub-v0.1; every sentence cites graph ids.

Contribution history

seq	event	detail
2285	APPENDED
2286	ACCEPT	AUTOMATIC_AFTER_VALIDATION
2453	LINK_EVIDENCE	by 03192ed8 (accepted)
3932	TASK_RESULT	by 03192ed8 (accepted)
7373	TASK_RESULT	by 8b5e81bf (accepted)
8072	TASK_RESULT	by 1efd24fe (accepted)
11921	LINK_EVIDENCE	by 1efd24fe (accepted)

Score trace

{
  "claim": "6bc79e76-e8e4-8d6d-bbb3-671d67896ddf",
  "links": [
    {
      "link": "09c7f98f-1c23-8179-8293-dcb64857eeda",
      "group": null,
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "2532988b-82c6-873e-94d7-08131e09f016",
      "direction": "SUPPORT",
      "magnitude": "1.800000",
      "relevance": "DIRECT",
      "extraction": "UNVERIFIED",
      "provenance": "INDEPENDENT",
      "observation": "DIRECT_TEXT",
      "source_type": "PRIMARY_TEXT",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "1.800000",
      "interpretive_steps": 0
    },
    {
      "kept": "6a355cdf-9b6e-86fa-a8f3-524fb81d0d22",
      "link": "3467073b-509e-8bca-9b76-04af301d6de6",
      "group": null,
      "reason": "dependent_strongest_only",
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "32c57b38-d467-839a-a046-eca996a715e7",
      "direction": "SUPPORT",
      "magnitude": "0.000000",
      "relevance": "DIRECT",
      "extraction": "UNVERIFIED",
      "provenance": "SELF",
      "observation": "DIRECT_TEXT",
      "source_type": "VIDEO",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "0.000000",
      "interpretive_steps": 1
    },
    {
      "kept": "6a355cdf-9b6e-86fa-a8f3-524fb81d0d22",
      "link": "673f1834-2ae0-816c-a124-7ac61523fe0c",
      "group": null,
      "reason": "dependent_strongest_only",
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "3358bd19-e978-82b6-9c8c-ea744a0e936e",
      "direction": "SUPPORT",
      "magnitude": "0.000000",
      "relevance": "MODERATE",
      "extraction": "UNVERIFIED",
      "provenance": "SELF",
      "observation": "HEARSAY",
      "source_type": "VIDEO",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "0.000000",
      "interpretive_steps": 1
    },
    {
      "link": "67c4e89c-c05b-8404-8dd1-87c294a3025b",
      "group": null,
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "62396ade-7a20-8ccb-8178-b859a0e4f9be",
      "direction": "SUPPORT",
      "magnitude": "0.459000",
      "relevance": "MODERATE",
      "extraction": "UNVERIFIED",
      "provenance": "INDEPENDENT",
      "observation": "DIRECT_TEXT",
      "source_type": "WEBSITE",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "0.459000",
      "interpretive_steps": 1
    },
    {
      "link": "6a355cdf-9b6e-86fa-a8f3-524fb81d0d22",
      "group": null,
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "6e7ed29a-00a1-8113-b984-02126bb7ee0c",
      "direction": "SUPPORT",
      "magnitude": "0.000000",
      "relevance": "WEAK",
      "extraction": "UNVERIFIED",
      "provenance": "SELF",
      "observation": "EXPERT_ANALYSIS",
      "source_type": "VIDEO",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "0.000000",
      "interpretive_steps": 1
    },
    {
      "link": "f5d4e085-dd4a-8294-8282-115747dc5f7e",
      "group": null,
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "408b75e1-f7a7-8388-b1dd-a248a3d67b44",
      "direction": "SUPPORT",
      "magnitude": "0.918000",
      "relevance": "STRONG",
      "extraction": "UNVERIFIED",
      "provenance": "INDEPENDENT",
      "observation": "DIRECT_TEXT",
      "source_type": "WEBSITE",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "0.918000",
      "interpretive_steps": 1
    }
  ],
  "model": "ledger-default@0.3.0",
  "prior": "0.50",
  "variants": {
    "spread": "0.0818",
    "permissive": "0.9842",
    "conservative": "0.9024"
  },
  "code_hash": "sha256:c61a6fdf35a754a0d7e958df149036a083c0eaeb7d2d2e52704ec32b0c36def7",
  "contested": false,
  "stability": "LOW",
  "claim_type": "INTERPRETIVE",
  "config_hash": "sha256:86768038080b8ab5233f1e4b7aba10af05b5bb44caa432095150b90e81209809",
  "probability": "0.9600",
  "provisional": true,
  "evidence_sum": "3.177000",
  "snapshot_seq": 11922,
  "prior_log_odds": "0.000000",
  "support_groups": 3,
  "model_dependent": true,
  "review_coverage": "0.75",
  "truth_evaluable": true,
  "assessment_state": "SUPPORTED",
  "review_checklist": {
    "qualifiers_reviewed": {
      "by": [
        "01a0bc53-0f71-797a-9ba0-f06f0fd572f9"
      ],
      "ok": true
    },
    "opposing_search_done": {
      "by": [
        "01a0bc53-0f5f-7497-91bd-78ee7847f851"
      ],
      "ok": true
    },
    "independence_reviewed": {
      "by": [],
      "ok": false
    },
    "primary_source_reviewed": {
      "by": [
        "09c7f98f-1c23-8179-8293-dcb64857eeda"
      ],
      "ok": true
    }
  },
  "contradict_groups": 0,
  "rounding_boundary": false,
  "posterior_log_odds": "3.177000",
  "not_applicable_reason": null,
  "independence_unreviewed": 6
}