Skip to content

Research

Complete evidence archives for every research instance. Every source scored, every search logged. Nothing summarized away.

How This Is Organized

Research Instance → a specific investigation (e.g., R0005). Tagged by subject matter. Multiple runs over time.

Run → a specific execution of the research. Timestamped, versioned by prompt and model. Runs within an instance are re-executions of the same research, not additions of new items. Each run is independent.

Entity → the individual items within a run: claims or queries. Each has a full evidence archive.

The drill-down: Instance → Run → Entity → Evidence (sources, searches, scorecards). Every level has a summary. Every level links to the next level of detail.

ID Namespace

Within a single claim or query, all entity IDs are unique — no prefix needed. Fully-qualified IDs (e.g., C001-SRC01-E01) appear only in page titles to establish position in the hierarchy. Everywhere else, use the short form.

Entity Short form Scope Notes
Hypothesis H1 Entity H1=affirmative, H2=negative, H3=nuanced
Search S01 Entity Unique within entity
Source SRC01 Entity Unique within entity
Search Result S01-R01 Entity Scoped to search
Evidence Extract SRC01-E01 Entity Scoped to source

Background Research

Query-driven investigations that inform articles. Each instance has one or more runs; the most recent run reflects the latest evidence.

R0065 — OSS CI Auditing: Adoption, Barriers, and Tooling

Tags: devops, technology · Runs: 2026-05-18 run-1, run-2, run-3 · 4 claims + 3 queries

40% CI adoption likely outdated. 5.4 Scorecard score misattributed (Wolfi, not 1M projects). OSSRA 87%/581/107% confirmed but M&A sample bias. 95% fix-available confirmed but 70% require breaking changes. Comprehensive CI gates well below 1% adoption.

A0030 — OSS CI Auditing Toolkit

devops, technology

R0048 — Corporate AI Training and Sycophancy Awareness

Tags: ai · Runs: 2026-03-29, 2026-04-01 · 3 queries

No corporate or government training warns about sycophancy under any name. 82% have AI training but it's superficial. 40% zero-scrutiny rate.

A0022 — AI Sycophancy: The Yes-Man Problem

ai

R0044 — Expanded Vocabulary Sycophancy Research

Tags: ai · Runs: 2026-03-29, 2026-04-01 · 4 queries

All regulations address system design but not system output. Design vs output gap confirmed.

A0022 — AI Sycophancy: The Yes-Man Problem

ai

R0043 — Sycophancy Vocabulary Mapping

Tags: ai · Runs: 2026-03-28, 2026-04-01 · 3 queries

System-side vs human-side vocabulary asymmetry. No shared vocabulary bridges them. Every bridging taxonomy omits sycophancy.

A0022 — AI Sycophancy: The Yes-Man Problem

ai

R0042 — Private AI Enterprise Motivations

Tags: ai · Runs: 2026-03-28, 2026-04-01 · 3 queries

Enterprise private AI motivated by data sovereignty, not behavioral customization. Anti-sycophancy treated as model provider's problem.

A0022 — AI Sycophancy: The Yes-Man Problem

ai

R0041 — Enterprise Sycophancy Reduction

Tags: ai · Runs: 2026-03-28, 2026-04-01 · 3 queries

No vendor offers enterprise anti-sycophancy products. RLVR limited to verifiable domains.

A0022 — AI Sycophancy: The Yes-Man Problem

ai

R0040 — RLHF Alternatives and Sycophancy Root Cause

Tags: ai · Runs: 2026-03-28, 2026-03-29, 2026-04-01 · 2 queries

Six+ RLHF alternatives mapped. Sycophancy root cause is biased preference data, not the RL algorithm.

A0022 — AI Sycophancy: The Yes-Man Problem

ai

R0051 — Fact-Checking Methodology Gap Analysis

Tags: methodology · Runs: 2026-03-31 · 3 queries

No formal epistemological framework in fact-checking comparable to GRADE, IPCC, or ICD 203.

A0012 — The Truth is Out There (Part 1)

methodology

R0050 — Journalism and Other Truth-Seeking Disciplines

Tags: methodology · Runs: 2026-03-31 · 3 queries

Journalism is principles-based, not methodology-based.

A0012 — The Truth is Out There (Part 1)

methodology

R0049 — Published AI Research Methodology Prompts

Tags: methodology · Runs: 2026-03-31 · 3 queries

No complete AI prompt implementing a full analytical rigor framework.

A0012 — The Truth is Out There (Part 1)

methodology

A0013 — The Truth is Out There (Part 2)

methodology

R0047 — Source-Back Verification Test

Tags: methodology · Runs: 2026-03-29 · 1 query

Validated source-back verification methodology. Caught two framing discrepancies that the process self-audit missed.

R0045 — 2001 OSCON Prediction Verification

Tags: technology, history · Runs: 2026-03-29 · 7 queries

Verifying predictions from the 2001 OSCON keynote.

A0024 — The Power of Prediction (draft)

technology

R0029 — AI Attribution Frameworks and Public Sentiment

Tags: ai · Runs: 2026-03-27 · 5 queries

AI attribution frameworks, public sentiment, journal AI authorship policies.

A0021 — On the Use of the Plural Voice

ai

R0027 — Multilingual Prompt Engineering Challenges

Tags: ai · Runs: 2026-03-26 · 3 queries

Cross-language performance gaps: 3-30pp. Vendor guides all English-only.

A0019 — Prompt Engineering Is Not. Engineering, That Is.

ai

R0026 — Pretendgineer Prior Art

Tags: ai · Runs: 2026-03-25 · 1 query

Earliest documented use: January 2009 (Urban Dictionary).

A0020 — Pretendgineer (draft)

ai

R0024 — Sycophancy, Addiction, and Vendor Incentives

Tags: ai · Runs: 2026-03-25 · 4 queries

Vendor disincentive to reduce sycophancy confirmed. "Dark addiction pattern" at CHI 2025.

A0019 — Prompt Engineering Is Not. Engineering, That Is.

ai

R0023 — Counterproductive Advice and Prompt Lifecycle

Tags: ai · Runs: 2026-03-25 · 4 queries

Expert personas degrade accuracy. GPT-4 dropped 84%→51% in 3 months.

A0019 — Prompt Engineering Is Not. Engineering, That Is.

ai

R0021 — Engineering Definitions and Standards

Tags: ai · Runs: 2026-03-25 · 8 queries

Formal engineering definitions, PE licensing, 84% subjective vendor guidance, 430:1 ambiguity gap.

A0019 — Prompt Engineering Is Not. Engineering, That Is.

ai

R0020 — Prompt Engineering Gaps

Tags: ai · Runs: 2026-03-25 · 4 queries

Testing frameworks, sycophancy in vendor docs, theory-practice gap.

A0019 — Prompt Engineering Is Not. Engineering, That Is.

ai

R0005 — AI Company Profitability Before 2030

Tags: ai · Runs: 2026-03-17 · 1 query

Company-by-company analysis across pure-play labs, diversified tech, and infrastructure providers.


Claim Verification

Blind fact-checking of published and draft articles. Claims are extracted from the article body and verified independently against evidence.

R0057 — A0022 Claim Verification (final)

Runs: 2026-04-01 · 33 claims · 100% pass rate

Final iteration. 5 Almost certain, 19 Very likely, 9 Likely.

A0022 — AI Sycophancy: The Yes-Man Problem

ai

R0056 — A0022 Claim Verification (iteration 2)

Runs: 2026-04-01 · 28 claims · 93% pass rate

2 corrections: unverifiable 84-85% figure, DeepSeek V3 details.

A0022 — AI Sycophancy: The Yes-Man Problem

ai

R0055 — A0022 Claim Verification (iteration 1)

Runs: 2026-04-01 · 28 claims · 93% pass rate

3 corrections + 5 softenings.

A0022 — AI Sycophancy: The Yes-Man Problem

ai

R0054 — A0013 Claim Verification (iteration 2)

Runs: 2026-03-31

A0013 — The Truth is Out There (Part 2)

methodology

R0053 — A0013 Claim Verification

Runs: 2026-03-31 · 7 claims

Choe uniqueness softened. Enforcement language nuanced.

A0013 — The Truth is Out There (Part 2)

methodology

R0052 — A0012 Claim Verification

Runs: 2026-03-31 · 14 claims

10 Almost certain, 4 Very likely. Blanket negatives softened.

A0012 — The Truth is Out There (Part 1)

methodology

R0031 — A0021 Claim Verification

Runs: 2026-03-27, 2026-03-29 · 14 claims

Two corrections. Rerun caught NeurIPS exception.

A0021 — On the Use of the Plural Voice

ai

R0028 — A0019 Claim Verification

Runs: 2026-03-26 · 33 claims

504 files. Five corrections. 100% source convergence with query research.

A0019 — Prompt Engineering Is Not. Engineering, That Is.

ai

R0007 — A0007 Claim Verification

Runs: 2026-03-19 · 15 claims

A0007 — AI Made Everyone Faster

ai

R0002 — Methodology Framework Verification

Runs: 2026-03-13 · 12 claims

Nine frameworks underlying the unified research methodology.

A0012 — The Truth is Out There (Part 1)

methodology

R9990 — STAR Interview Neurodivergent Impact

Runs: 2026-03-20 · 1 claim

Does the STAR format disadvantage neurodivergent individuals?