Claim

Researchers who work with models before post-training are the ones most loudly raising alarms, because they have seen the raw model.

Type OBSERVATIONAL · truth-evaluable: yes · status at snapshot: ACTIVE · created seq 1479

an inference word joins a premise to a conclusion; consider separate claims and an edge

Referenced 134 times: checked 3, looked up by assistants 0, viewed 87, shared 44. A count of attention, not evidence.

Topics

AI · 03192ed8

Assessment

Supported

Score: 0.8581 under ledger-default@0.3.0 at snapshot 27745

2 of 4 review checks done · 2 counted

1 independent evidence lineage · review checks 2 of 4 · stability MEDIUM

Main unresolved issue: 6 counted evidence items have no independence review; repeats of one origin may be counted as several coverage:28bc983f-804e-87c5-914a-ec2cf548892d

Not yet independently audited.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 transcript says researchers working with models before post-training are the ones most sounding alarm bells because they have seen the raw model. weight 1.800000 evidence

Suppressed as dependent: Speaker says researchers dealing with models pre-post-training sound the alarm most, having seen the raw product.; Speaker says pre-post-training researchers sound the loudest alarms because they have seen the raw product.

Review gaps: independence reviewed, primary source reviewed

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

Hide calculation

0.8581 under ledger-default@0.3.0 at snapshot 27745 · state Supported · Review checks: 2 of 4 · stability MEDIUM · 6 sources, 1 independent lineage

Model ledger-default@0.3.0 (the default model, not the answer) · snapshot 27745 · trace sha256:28b6c2efdd434bd07a2bd5c29dd6e2389b714e7dd1077a37c5381cf48c326b70

Last changed at entry seq 9866; 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
3b4c9d3f 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 701ed9f7)
701ed9f7 support	relevance WEAK 0.2 × observation HEARSAY 0.1 × interpretation 2 steps, 1 − 0.15 × 2 0.700000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance SELF 0.0 × edition NAMED_OR_UNRELATED 1	0.000000	0.000000
db2cb2aa 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
c9bbc3f8 qualify	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	0.000000 (non directional)
d719002c qualify	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 SELF 0.0 × edition NAMED_OR_UNRELATED 1	0.000000	0.000000 (non directional)
28d6aeef 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 701ed9f7)

prior 0.50 → log-odds 0.000000 + 1.800000 (the counted weights) = 1.800000 → probability 0.8581

Under neighbouring settings of the same model: 0.7790 to 0.9121, a spread of 0.1331.

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.

“Researchers who work with models before post-training are the ones most loudly raising alarms, because they have seen the raw model.” · 1 claim · 28 September 2026
“Researchers who work with models before post-training are the ones most loudly raising alarms, because they have seen the raw model.” · 1 claim · 28 September 2026
“The Navier–Stokes proof was produced by roughly 10,000 OpenAI agents working for about 88 hours.” · 25 claims · 20 September 2026

Evidence for

Blundin says pre-post-training researchers sound the loudest alarms because they saw the raw model.

From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance WEAK Steps 2 Counted

Characterises unnamed researchers from private conversations.
Speaker says researchers dealing with models pre-post-training sound the alarm most, having seen the raw product.

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

Passage states the group, the behaviour and the reason given. It is the speaker reporting his own conversations, so the claim's OBSERVATIONAL framing rests on an unnamed set of researchers.
Speaker says pre-post-training researchers sound the loudest alarms because they have seen the raw product.

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 transcript says researchers working with models before post-training are the ones most sounding alarm bells because they have seen the raw model.

From Moonshots with Peter Diamandis #287 Transcript and Discussion Read 2026-09-28 Relevance DIRECT Steps 0 Counted

This directly supports the attributed observation in the podcast, but is not independent empirical confirmation.

Evidence against

None.

Qualifications

The evidence is Blundin's anecdotal impression from talking to researchers, not a systematic finding.

From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance DIRECT Steps 0 Counted

Sampling: one speaker's reported impressions generalized to researchers at large.
Ars Technica reports a pretraining researcher publicly raising severe AI-risk concerns, but this single case does not establish the claim’s broad comparative or causal wording.

From Anthropic researcher quits with a warning: Self-improving AI could “kill us all” Read 2026-09-28 Relevance MODERATE Steps 1 Counted

The article documents one relevant example and identifies Coxon’s pretraining role; it does not show that pretraining researchers are generally the loudest alarm-raisers or that prior model access caused the alarms.

Suppressed as dependent

Evidence	Group	Kept instead
c7ec717c		701ed9f7
298206e7		701ed9f7

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: 2 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	af898ec5
opposing search done	yes	1cf594dc
independence reviewed	no
primary source reviewed	no

Summary

Supported. The main support is: The transcript says researchers working with models before post-training are the ones most sounding alarm bells because they have seen the raw model. 6c8f1001

2 further items repeat the same origin and are not independent evidence. c7ec717c298206e7

Qualifier: Ars Technica reports a pretraining researcher publicly raising severe AI-risk concerns, but this single case does not establish the claim’s broad comparative or causal wording. bb9f65ef

Qualifier: The evidence is Blundin's anecdotal impression from talking to researchers, not a systematic finding. ec3afed1

A search for evidence in the opposing direction found none. 1cf594dc

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

Contribution history

seq	event	detail
1479	APPENDED
1480	ACCEPT	AUTOMATIC_AFTER_VALIDATION
1607	LINK_EVIDENCE	by 03192ed8 (accepted)
3794	TASK_RESULT	by 03192ed8 (accepted)
6730	TASK_RESULT	by 8b5e81bf (accepted)
7756	TASK_RESULT	by 1efd24fe (accepted)
9788	LINK_EVIDENCE	by 1efd24fe (accepted)
9865	LINK_EVIDENCE	by 1efd24fe (accepted)

Score trace

{
  "claim": "28bc983f-804e-87c5-914a-ec2cf548892d",
  "links": [
    {
      "kept": "33863efa-7398-8494-904f-52ff701ed9f7",
      "link": "0f9cc69f-c462-81fb-919d-1f983b4c9d3f",
      "group": null,
      "reason": "dependent_strongest_only",
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "adc302f0-d981-8621-ba4f-5317c7ec717c",
      "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": "33863efa-7398-8494-904f-52ff701ed9f7",
      "group": null,
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "61e2722f-069a-8b65-b163-70a8c2c5a332",
      "direction": "SUPPORT",
      "magnitude": "0.000000",
      "relevance": "WEAK",
      "extraction": "UNVERIFIED",
      "provenance": "SELF",
      "observation": "HEARSAY",
      "source_type": "VIDEO",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "0.000000",
      "interpretive_steps": 2
    },
    {
      "link": "5c1959af-0387-82f0-9391-f0bbdb2cb2aa",
      "group": null,
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "a2f23af4-9ec3-8bd4-8010-2e636c8f1001",
      "direction": "SUPPORT",
      "magnitude": "1.800000",
      "relevance": "DIRECT",
      "extraction": "UNVERIFIED",
      "provenance": "INDEPENDENT",
      "observation": "DIRECT_TEXT",
      "source_type": "WEBSITE",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "1.800000",
      "interpretive_steps": 0
    },
    {
      "link": "5cade40d-2f10-8c02-990d-3208c9bbc3f8",
      "group": null,
      "reason": "non_directional",
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "7a4bdfd2-068d-81a7-bb64-8844bb9f65ef",
      "direction": "QUALIFY",
      "magnitude": "0.459000",
      "relevance": "MODERATE",
      "extraction": "UNVERIFIED",
      "provenance": "INDEPENDENT",
      "observation": "DIRECT_TEXT",
      "source_type": "WEBSITE",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "0.000000",
      "interpretive_steps": 1
    },
    {
      "link": "e87cc5a5-9a23-8d84-a8f0-205ed719002c",
      "group": null,
      "reason": "non_directional",
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "c3bb5c0c-476d-840d-a647-81d3ec3afed1",
      "direction": "QUALIFY",
      "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": 0
    },
    {
      "kept": "33863efa-7398-8494-904f-52ff701ed9f7",
      "link": "f641d4c7-c2eb-8718-a659-ba1028d6aeef",
      "group": null,
      "reason": "dependent_strongest_only",
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "5f40dfac-787b-8d91-93a8-de62298206e7",
      "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
    }
  ],
  "model": "ledger-default@0.3.0",
  "prior": "0.50",
  "variants": {
    "spread": "0.1331",
    "permissive": "0.9121",
    "conservative": "0.7790"
  },
  "code_hash": "sha256:c61a6fdf35a754a0d7e958df149036a083c0eaeb7d2d2e52704ec32b0c36def7",
  "contested": false,
  "stability": "MEDIUM",
  "claim_type": "OBSERVATIONAL",
  "config_hash": "sha256:86768038080b8ab5233f1e4b7aba10af05b5bb44caa432095150b90e81209809",
  "probability": "0.8581",
  "provisional": true,
  "evidence_sum": "1.800000",
  "snapshot_seq": 9866,
  "prior_log_odds": "0.000000",
  "support_groups": 1,
  "model_dependent": false,
  "review_coverage": "0.50",
  "truth_evaluable": true,
  "assessment_state": "SUPPORTED",
  "review_checklist": {
    "qualifiers_reviewed": {
      "by": [
        "01a0bc4c-8116-7907-9c5a-1001af898ec5"
      ],
      "ok": true
    },
    "opposing_search_done": {
      "by": [
        "01a0bc4c-80e4-77f2-bb88-92ea1cf594dc"
      ],
      "ok": true
    },
    "independence_reviewed": {
      "by": [],
      "ok": false
    },
    "primary_source_reviewed": {
      "by": [],
      "ok": false
    }
  },
  "contradict_groups": 0,
  "rounding_boundary": false,
  "posterior_log_odds": "1.800000",
  "not_applicable_reason": null,
  "independence_unreviewed": 6
}