starintel-labs/starintel-gpt-auto-dig

Fork 0

[GH-#68] Seed dataset: DOGE, USAID cuts, mortality estimates, and Elon Musk accountability claims #998

New issue

Open

opened 2026-09-08 17:05:44 +00:00 by nsaspy · 0 comments

nsaspy commented

2026-09-08 17:05:44 +00:00

Owner

Mirrored from GitHub https://github.com/lost-rob0t/starintel-gpt-auto-dig/issues/68 (GitHub is authoritative for this item).

Objective

Create a validated StarIntel v0.9.0 research packet and topic dataset covering the dismantling/defunding of USAID, the role attributed to Elon Musk and DOGE, modeled mortality estimates, individually reported deaths, and competing political/accountability claims.

Primary seed

GDF, “Elon Musk, Mass Murderer?”
https://www.youtube.com/watch?v=lB5Sg10_gKM

Treat the video as a secondary narrative and lead-generation source. Do not use its title or commentary as proof that Musk committed murder.

Proposed dataset

usaid-doge-mortality-accountability

Suggested run path:

digs/usaid-doge-mortality-accountability/2026-07-28-gdf-video-seed/

Required evidence layers

Keep these as distinct document classes and graph branches:

Observed government actions
- foreign-aid pause, award terminations, agency restructuring/dissolution, transfer of remaining functions, payment interruptions, program closures;
- identify the formal decision-maker, legal authority, implementation chain, dates, agencies, and contractors.
Modeled mortality estimates
- never store projections or counters as confirmed death tallies;
- preserve methodology, baseline, counterfactual, date range, confidence/uncertainty, diseases/programs included, funding assumptions, and model revisions.
Individually documented cases
- create separate event/claim/source records for named deaths reported by journalists or health workers;
- distinguish direct documentation from causal inference about whether continued USAID support would have prevented the death.
Statements and rhetoric
- preserve exact speaker and source for claims such as “zero died,” “got carried away,” “mass murderer,” criminal accountability, and demands for investigation;
- encode these as attributed claims, not neutral facts.

High-priority sources

Primary research

Cavalcanti et al., The Lancet (2025), “Evaluating the impact of two decades of USAID interventions and projecting the effects of defunding on mortality up to 2030”
- DOI: https://doi.org/10.1016/S0140-6736(25)01186-9
- PubMed: https://pubmed.ncbi.nlm.nih.gov/40609560/
- Full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC12274115/

Current mortality model/counter

Boston University School of Public Health, Brooke Nichols / Impact Counter background
- https://www.bu.edu/sph/news/articles/2025/tracking-anticipated-deaths-from-usaid-funding-cuts/

Named-case reporting

NPR/KPBS, “Trump's team says 'no children' died from USAID cuts. Consider these 3 cases”
- https://www.kpbs.org/news/health/2026/07/17/trumps-team-says-no-children-died-from-usaid-cuts-consider-these-3-cases

Musk statements and interview

The Economist interview with Elon Musk, July 2026; obtain the canonical interview/video/transcript.
Verify and source separately:
- “I think I got a little too involved in politics, got carried away, frankly.”
- Musk’s claim that USAID/DOGE cuts caused “zero” deaths.

Government and legal record

Collect official executive orders, White House/State Department releases, congressional material, court filings, inspector-general reports, contract termination records, appropriations/outlay data, and archived USAID pages.

Initial entities

People

Elon Musk
Donald Trump
Marco Rubio
Ro Khanna
Brooke Nichols
Daniella Medeiros Cavalcanti
Davide Rasella
James Macinko
Zanny Minton Beddoes
GDF / video presenter(s), once identified

Organizations

United States Agency for International Development (USAID)
Department of Government Efficiency (DOGE)
U.S. Department of State
White House
U.S. Congress
The Lancet
Boston University School of Public Health
UCLA Fielding School of Public Health
Barcelona Institute for Global Health (ISGlobal)
The Economist
NPR
GDF

Core claims to represent separately

Continued deep USAID cuts could produce more than 14 million additional deaths by 2030, including roughly 4.5 million children under five.
USAID-supported programs were associated with large historical mortality reductions and an estimated 91 million deaths prevented over 2001–2021.
The Impact Counter estimates more than 700,000 deaths over roughly the first year following abrupt aid reductions; this is a model-derived estimate, not a registry of confirmed individual deaths.
Musk acknowledged becoming overly involved in politics and said he “got carried away.”
Musk denied that the cuts caused deaths.
Critics, including political figures and commentators, characterize Musk’s role using criminal or mass-killing language.

Every claim must include exact wording where available, speaker/author, publication date, retrieval date, source URI, model/claim type, and verification state.

Research questions

What authority did Musk and DOGE formally possess over USAID decisions?
Which actions were ordered by Trump, Rubio, Musk, DOGE personnel, OMB, State Department, or agency officials?
Which contracts, payments, disease programs, food programs, and field operations were stopped, restored, waived, or transferred?
What are the Lancet study’s assumptions, model design, sensitivity analyses, limitations, and strongest published critiques?
How does the Impact Counter differ methodologically from the Lancet forecast?
Which deaths are individually documented, and what evidence supports causal attribution to interrupted aid?
Which numbers are historical estimates, forecasts, real-time modeled counters, reported cases, or political talking points?
What legal theories could apply to decision-makers, and how do they differ from moral or rhetorical responsibility?

Required output

README.md
sources.md
canonical starintel-documents.jsonl
manifest.json
dataset-manifest document
source documents for every cited item
separate person/org/event/claim/relation/research-pass documents
explicit relations connecting policy actions → program interruptions → modeled or observed outcomes
validation through python3 scripts/validate-for-merge.py --site
draft PR only; do not mark ready or merge until the full local gate and all GitHub checks pass

Guardrails

Do not collapse correlation, modeling, projection, attribution, and adjudicated fact.
Do not record “mass murderer” as a verified identity or legal conclusion.
Do not treat modeled cumulative deaths as individually confirmed victims.
Preserve contrary evidence, uncertainty, methodological criticism, and revisions.
Do not infer that every USAID interruption was personally ordered by Musk.
Use exact dates and authoritative records wherever available.

Mirrored from GitHub https://github.com/lost-rob0t/starintel-gpt-auto-dig/issues/68 (GitHub is authoritative for this item). --- ## Objective Create a validated StarIntel v0.9.0 research packet and topic dataset covering the dismantling/defunding of USAID, the role attributed to Elon Musk and DOGE, modeled mortality estimates, individually reported deaths, and competing political/accountability claims. ## Primary seed - GDF, **“Elon Musk, Mass Murderer?”** - https://www.youtube.com/watch?v=lB5Sg10_gKM Treat the video as a **secondary narrative and lead-generation source**. Do not use its title or commentary as proof that Musk committed murder. ## Proposed dataset `usaid-doge-mortality-accountability` Suggested run path: `digs/usaid-doge-mortality-accountability/2026-07-28-gdf-video-seed/` ## Required evidence layers Keep these as distinct document classes and graph branches: 1. **Observed government actions** - foreign-aid pause, award terminations, agency restructuring/dissolution, transfer of remaining functions, payment interruptions, program closures; - identify the formal decision-maker, legal authority, implementation chain, dates, agencies, and contractors. 2. **Modeled mortality estimates** - never store projections or counters as confirmed death tallies; - preserve methodology, baseline, counterfactual, date range, confidence/uncertainty, diseases/programs included, funding assumptions, and model revisions. 3. **Individually documented cases** - create separate event/claim/source records for named deaths reported by journalists or health workers; - distinguish direct documentation from causal inference about whether continued USAID support would have prevented the death. 4. **Statements and rhetoric** - preserve exact speaker and source for claims such as “zero died,” “got carried away,” “mass murderer,” criminal accountability, and demands for investigation; - encode these as attributed claims, not neutral facts. ## High-priority sources ### Primary research - Cavalcanti et al., *The Lancet* (2025), “Evaluating the impact of two decades of USAID interventions and projecting the effects of defunding on mortality up to 2030” - DOI: https://doi.org/10.1016/S0140-6736(25)01186-9 - PubMed: https://pubmed.ncbi.nlm.nih.gov/40609560/ - Full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC12274115/ ### Current mortality model/counter - Boston University School of Public Health, Brooke Nichols / Impact Counter background - https://www.bu.edu/sph/news/articles/2025/tracking-anticipated-deaths-from-usaid-funding-cuts/ ### Named-case reporting - NPR/KPBS, “Trump's team says 'no children' died from USAID cuts. Consider these 3 cases” - https://www.kpbs.org/news/health/2026/07/17/trumps-team-says-no-children-died-from-usaid-cuts-consider-these-3-cases ### Musk statements and interview - The Economist interview with Elon Musk, July 2026; obtain the canonical interview/video/transcript. - Verify and source separately: - “I think I got a little too involved in politics, got carried away, frankly.” - Musk’s claim that USAID/DOGE cuts caused “zero” deaths. ### Government and legal record Collect official executive orders, White House/State Department releases, congressional material, court filings, inspector-general reports, contract termination records, appropriations/outlay data, and archived USAID pages. ## Initial entities ### People - Elon Musk - Donald Trump - Marco Rubio - Ro Khanna - Brooke Nichols - Daniella Medeiros Cavalcanti - Davide Rasella - James Macinko - Zanny Minton Beddoes - GDF / video presenter(s), once identified ### Organizations - United States Agency for International Development (USAID) - Department of Government Efficiency (DOGE) - U.S. Department of State - White House - U.S. Congress - The Lancet - Boston University School of Public Health - UCLA Fielding School of Public Health - Barcelona Institute for Global Health (ISGlobal) - The Economist - NPR - GDF ## Core claims to represent separately - Continued deep USAID cuts could produce more than 14 million additional deaths by 2030, including roughly 4.5 million children under five. - USAID-supported programs were associated with large historical mortality reductions and an estimated 91 million deaths prevented over 2001–2021. - The Impact Counter estimates more than 700,000 deaths over roughly the first year following abrupt aid reductions; this is a model-derived estimate, not a registry of confirmed individual deaths. - Musk acknowledged becoming overly involved in politics and said he “got carried away.” - Musk denied that the cuts caused deaths. - Critics, including political figures and commentators, characterize Musk’s role using criminal or mass-killing language. Every claim must include exact wording where available, speaker/author, publication date, retrieval date, source URI, model/claim type, and verification state. ## Research questions 1. What authority did Musk and DOGE formally possess over USAID decisions? 2. Which actions were ordered by Trump, Rubio, Musk, DOGE personnel, OMB, State Department, or agency officials? 3. Which contracts, payments, disease programs, food programs, and field operations were stopped, restored, waived, or transferred? 4. What are the Lancet study’s assumptions, model design, sensitivity analyses, limitations, and strongest published critiques? 5. How does the Impact Counter differ methodologically from the Lancet forecast? 6. Which deaths are individually documented, and what evidence supports causal attribution to interrupted aid? 7. Which numbers are historical estimates, forecasts, real-time modeled counters, reported cases, or political talking points? 8. What legal theories could apply to decision-makers, and how do they differ from moral or rhetorical responsibility? ## Required output - `README.md` - `sources.md` - canonical `starintel-documents.jsonl` - `manifest.json` - dataset-manifest document - source documents for every cited item - separate person/org/event/claim/relation/research-pass documents - explicit relations connecting policy actions → program interruptions → modeled or observed outcomes - validation through `python3 scripts/validate-for-merge.py --site` - draft PR only; do not mark ready or merge until the full local gate and all GitHub checks pass ## Guardrails - Do not collapse correlation, modeling, projection, attribution, and adjudicated fact. - Do not record “mass murderer” as a verified identity or legal conclusion. - Do not treat modeled cumulative deaths as individually confirmed victims. - Preserve contrary evidence, uncertainty, methodological criticism, and revisions. - Do not infer that every USAID interruption was personally ordered by Musk. - Use exact dates and authoritative records wherever available.

nsaspy added the

github-mirror

label

2026-09-08 17:05:44 +00:00

No labels

github-mirror

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

starintel-labs/starintel-gpt-auto-dig#998

No description provided.

Rows
Columns