← Cockpit
231_017predictionAIAI-scaling

A major supply chain attack from untrusted open-weight code generation models is absolutely a threat vector.

Predictor: Alex Wissner-Gross · ep#231 "Top AI News: Sonnet 4.6, Grok 4.2, Gemini 3 Deep Think, and OpenClaw | EP #231" · source

Prior probability
50.0%
Current probability
41.9%
evolves via intake + LBP
Conviction
4/5
Signal quality
B
Resolution
pending
Window
2028-06-01 – 2028-06-30
Edges in / out
10 / 5
Tickers exposed
37

Prediction text

A major supply chain attack from untrusted open-weight code generation models is absolutely a threat vector. | the world to my knowledge has not seen a major supply chain attack yet. that stems from the result of untrusted openweight code generation models rewriting the entire supply chain. But do I think that's possible? Yes, I think that is absolutely a threat vector.

Verbatim quote

From episode "Top AI News: Sonnet 4.6, Grok 4.2, Gemini 3 Deep Think, and OpenClaw | EP #231"
the world to my knowledge has not seen a major supply chain attack yet. that stems from the result of untrusted openweight code generation models rewriting the entire supply chain. But do I think that's possible? Yes, I think that is absolutely a threat vector.

Predictor: Alex Wissner-Gross

κ + Brier as of 2026-05-22
κ (discount)
0.844
Brier
0.0341
excellent
Hits / Misses
6 / 1
of 11 resolved
Hit rate
54.5%
Calibration plot (stated vs observed)

Evidence about this node from Alex Wissner-Gross is multiplied by κ in /api/intake. Lower κ = less weight; floors at 0.10 (effectively silenced) and caps at 1.00 (full weight).

Reference class

Not linked

This node isn't linked to a reference class. The Bayesian update applies without outside-view blending.

Probability over time

4 prob_history rows
0%25%50%75%100%prior 50%2026-04-302026-05-032026-05-10
intake v2milestone miss sweeplbp propagationreference class assignedlegacy v1prior_prob (analyst seed)current = 41.9%

Milestone chain

Pre-event signals (upstream prereqs + window checkpoints) → resolution event → downstream cascades. Status/dates update from linked nodes; re-derive nightly via scripts/ops/derive_milestones.py.
Leading chain: 8 fired ✓ · 2 pending
  1. 2025-09-15hitShai-Hulud npm self-replicating worm with LLM-generated bash script (Sept 2025)
    How: ReversingLabs detects self-replicating worm on npm registry with bash script bearing telltale LLM-generation comment/emoji signatures
    Source: https://asecurityengineer.com/posts/the-rise-of-supply-chain-attacks/ — Shai-Hulud Sept 15 2025 inflection pointconf 95%
    Notes: HIT — first publicly attributed LLM-authored worm in npm registry. Direct precursor to Wissner-Gross threat-vector claim.
  2. 2025-07-01hitAmazon Q VS Code extension compromised — wipe-files prompt injected
    How: Verified open-weight/agentic AI tool ships compromised version with malicious prompt that passes vendor verification and reaches end users
    Source: https://fortune.com/2025/12/15/ai-coding-tools-security-exploit-software/conf 90%
    Notes: HIT — Amazon Q extension compromise distributed file-deletion prompt to users. Attack vector validated.
  3. 2026-03-31hitLiteLLM PyPI supply chain compromise exposes Mercor AI training data
    How: OECD AI Incidents Tracker logs supply chain attack on open-source LLM library on PyPI exposing customer data
    Source: https://oecd.ai/en/incidents/2026-03-31-e055 — LiteLLM PyPI compromise (~3.4M daily downloads, two malicious versions live ~3 hours)conf 95%
    Notes: HIT — confirms LLM-tooling supply chain as live attack vector. Threat vector is no longer hypothetical.
  4. 2026-06-01 → 2027-12-31pendingFirst mass-impact (>100K orgs) supply chain attack with confirmed LLM-rewritten dependency
    How: Single attack chain causes >100,000 confirmed downstream organizations compromised, with forensic attribution of LLM-driven code rewrite as the entry vector
    Source: OECD AI Incidents Tracker, CISA advisoriesconf 60%
    Notes: Trajectory escalates from Shai-Hulud (npm worm) and LiteLLM (PyPI). Open question is when scale crosses regulatory threshold.
  5. 2026-06-01 → 2028-06-01pendingLoRA adapter or model-weight backdoor attack publicly attributed in Hugging Face / public registry
    How: First publicly disclosed incident where a malicious LoRA adapter or open-weight model checkpoint introduces a backdoor used in subsequent supply chain compromise
    Source: Hugging Face security disclosures, NeuralTrust 2026 supply chain reportsconf 70%
    Notes: NullBulge already weaponized Hugging Face/GitHub repos. LoRA-level backdoors flagged as next-generation vector.

What if this resolves?

Clamp this prediction TRUE or FALSE and run a counterfactual Gibbs sample. Surfaces the predictions whose marginals shift most under that assumption.
(live posterior: 42%)

Click a button to clamp this prediction and run a Gibbs sample. Returns the predictions whose marginals shift most. ~30s per run; ideal for stress-testing "if X resolves, what else moves?"

Evidence chain

Every probability update with full Bayesian provenance — chronological, latest first
LBP2026-05-10T02:00:02Z41.9%-1.0pp
Network propagation: 42.9% → 41.9%
6-iter LBP, residual 0.00584 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run e5c18d29
LBP2026-05-03T02:00:01Z42.9%-1.6pp
Network propagation: 44.5% → 42.9%
6-iter LBP, residual 0.00677 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run 1a683ac9
LBP2026-04-30T16:39:51Z44.5%-2.4pp
Network propagation: 46.8% → 44.5%
5-iter LBP, residual 0.00825 · damping 0.5, w_intrinsic 0.5 · method lbp_v2 · run 0c8a4ea3
LBP2026-04-30T02:18:57Z46.8%-3.2pp
Network propagation: 50.0% → 46.8%
5-iter LBP, residual 0.00825 · damping 0.5, w_intrinsic 0.5 · method lbp_v1 · run 592311ef

Network propagation neighbors

Top edges sorted by latest LBP cross-impact
All propagation →

Top incoming (parents)

Edges that influence THIS node's belief

KindNodeTheir probP(c|s=T)P(c|s=F)Δ implied
killerTK09
Energy Grid Cap (Data Center Power Wall)
35.0%0.0500.500-0.077
prereqSEM_027
Nvidia Data Center revenue +66% YoY, contributing ~90% of $5Joseph Moore
68.3%0.5000.050-0.067
killerTK05
Rate Regime Persistence (10y > 5% through 2028)
30.0%0.0500.500-0.054
killerTK03
AI Regulatory Moratorium (EU/US Capability Freeze)
10.0%0.0500.500+0.036
prereqSEM_012
Nvidia quadrupled chip production output while only doublingJensen Huang
75.0%0.5000.050-0.035

Top outgoing (children)

Predictions THIS node influences

KindNodeTheir probP(c|s=T)P(c|s=F)Δ implied
prereq246_017
Europa Clipper will arrive at Jupiter in 2030, conducting 50Peter Diamandis
37.7%0.6500.050-0.080
prereq247_035
Dario Amodei will solve most/all neurological diseases by enDario Amodei
38.8%0.7000.050-0.070
prereq246_016
Dragonfly nuclear-powered octicopter arrives at Titan in 203Peter Diamandis
35.6%0.6500.050-0.059
prereq230_020
Peter's 14-year-old son Milan will never get a driver's licePeter Diamandis
34.7%0.6500.050-0.049
prereq232_055
We're exiting the industrial age permanently as recursive sePeter Diamandis
35.5%0.7000.050-0.037

Ticker exposure

37 ticker(s) linked

Beneficiaries (24)

MUWULFIRENEQIXALABAPLDASMIYASMLPLABNVDANBISCRWVAAPLAMTAMZNDELLGOOGLIRMLNVGYMETAMSFTORCLSFTBYSTX

Adverse (6)

ACNGENCHGGIBMWNSLRN

Prerequisites (10)

Predictions that must hit first
TypePredTitleDomainLag
prereqSEM_011Nvidia became the world's first $5 trillion company (late 2025), operating a near-monopoly on advanced AI chips.Capital Markets
prereqSEM_027Nvidia Data Center revenue +66% YoY, contributing ~90% of $57B fiscal Q3 revenue; >$4.5T market cap entirely underpinned by AI silicon.Capital Markets
prereqSEM_014Nvidia's Arizona-based TSMC factory successfully fabricated cutting-edge semiconductors on US soil for first time in decades (October 2025).Manufacturing
prereqSEM_029Blackwell RTX PRO 5000 (72GB) engineered with 50% memory boost over previous generation — deliberate architectural concession for larger AI training.Semis/Products
prereqSEM_012Nvidia quadrupled chip production output while only doubling human headcount — achieved by deploying AI coding tools (Cursor, Claude Code) across engineering.AI/Manufacturing
killerTK09Energy Grid Cap (Data Center Power Wall)
killerTK05Rate Regime Persistence (10y > 5% through 2028)
killerTK01AGI Capability Plateau (2026-27 Training Stall)
killerTK02AI Compute Supply Shock (TSMC/Taiwan Disruption)
killerTK03AI Regulatory Moratorium (EU/US Capability Freeze)

Dependents (5)

Predictions enabled by this
TypePredTitleDomainLag
prereq232_055We're exiting the industrial age permanently as recursive self-improvement unfolds.AI
prereq247_035Dario Amodei will solve most/all neurological diseases by end of decadeBiotech/Longevity
prereq230_020Peter's 14-year-old son Milan will never get a driver's license.Auto/Transport
prereq246_017Europa Clipper will arrive at Jupiter in 2030, conducting 50 passes near Europa.Space
prereq246_016Dragonfly nuclear-powered octicopter arrives at Titan in 2034.Space

Linked documents (10)

Auto-generated by cosine similarity from Polymarket / Manifold / EDGAR / GDELT
SimSourceTitleMarket probPolarityReviewedPublished
0.645arxivDiffusionHijack: Supply-Chain PRNG Backdoor Attack on Diffusion Models and Quantum Random Number Defensementionspending2026-05-13
0.639arxivSteering Vectors are an Adversarial Attack Surfacementionspending2026-06-04
0.628arxivVectorSmuggle: Steganographic Exfiltration in Embedding Stores and a Cryptographic Provenance Defensementionspending2026-05-13
0.624arxivAttack Detection using Time Series Foundation Modelsmentionspending2026-06-04
0.616arxivCode-Centric Detection of Vulnerability-Fixing Commits: A Unified Benchmark and Empirical Studymentionspending2026-05-13
0.614arxivOn the Security of Research Artifactsmentionspending2026-05-07
0.608arxivLeakDojo: Decoding the Leakage Threats of RAG Systemsmentionspending2026-05-07
0.585github_releasetensorflow/tensorflow v2.18.1mentionspending2025-03-11
0.582github_releasetensorflow/tensorflow v2.14.1mentionspending2023-11-14
0.578manifoldfigure out the code - Insider trading and bluffingmentionspending2026-05-06

Raw metadata

From Thesis_Timeline_v1.0_FINAL workbook
{
  "nia": false,
  "url": "https://www.youtube.com/watch?v=HklyjXKYFng",
  "mode": "PREDICTION",
  "role": "Host",
  "context": "the world to my knowledge has not seen a major supply chain attack yet. that stems from the result of untrusted openweight code generation models rewriting the entire supply chain. But do I think that's possible? Yes, I think that is absolutely a threat vector.",
  "to_year": 2030,
  "verbatim": "the world to my knowledge has not seen a major supply chain attack yet. that stems from the result of untrusted openweight code generation models rewriting the entire supply chain. But do I think that's possible? Yes, I think that is absolutely a threat vector.",
  "conv_cues": "absolutely",
  "direction": "HAPPEN",
  "from_year": 2026,
  "timeframe": "future",
  "conv_level": "HIGH",
  "milestones": [
    {
      "kind": "llm_pre_event",
      "label": "Shai-Hulud npm self-replicating worm with LLM-generated bash script (Sept 2025)",
      "notes": "HIT — first publicly attributed LLM-authored worm in npm registry. Direct precursor to Wissner-Gross threat-vector claim.",
      "source": "https://asecurityengineer.com/posts/the-rise-of-supply-chain-attacks/ — Shai-Hulud Sept 15 2025 inflection point",
      "status": "hit",
      "weight": 0.4,
      "ordinal": -10,
      "source_id": null,
      "confidence": 0.95,
      "source_url": "https://asecurityengineer.com/posts/the-rise-of-supply-chain-attacks/",
      "expected_date": "2025-09-15",
      "observed_date": "2025-09-15",
      "research_origin": "deep_research",
      "measurement_criterion": "ReversingLabs detects self-replicating worm on npm registry with bash script bearing telltale LLM-generation comment/emoji signatures"
    },
    {
      "kind": "llm_pre_event",
      "label": "Amazon Q VS Code extension compromised — wipe-files prompt injected",
      "notes": "HIT — Amazon Q extension compromise distributed file-deletion prompt to users. Attack vector validated.",
      "source": "https://fortune.com/2025/12/15/ai-coding-tools-security-exploit-software/",
      "status": "hit",
      "weight": 0.4,
      "ordinal": -9,
      "source_id": null,
      "confidence": 0.9,
      "source_url": "https://fortune.com/2025/12/15/ai-coding-tools-security-exploit-software/",
      "expected_date": "2025-12-15",
      "observed_date": "2025-07-01",
      "research_origin": "deep_research",
      "measurement_criterion": "Verified open-weight/agentic AI tool ships compromised version with malicious prompt that passes vendor verification and reaches end users"
    },
    {
      "kind": "llm_pre_event",
      "label": "LiteLLM PyPI supply chain compromise exposes Mercor AI training data",
      "notes": "HIT — confirms LLM-tooling supply chain as live attack vector. Threat vector is no longer hypothetical.",
      "source": "https://oecd.ai/en/incidents/2026-03-31-e055 — LiteLLM PyPI compromise (~3.4M daily downloads, two malicious versions live ~3 hours)",
      "status": "hit",
      "weight": 0.4,
      "ordinal": -8,
      "source_id": null,
      "confidence": 0.95,
      "source_url": "https://oecd.ai/en/incidents/2026-03-31-e055",
      "expected_date": "2026-03-31",
      "observed_date": "2026-03-31",
      "research_origin": "deep_research",
      "measurement_criterion": "OECD AI Incidents Tracker logs supply chain attack on open-source LLM library on PyPI exposing customer data"
    },
    {
      "kind": "prereq",
      "label": "Nvidia became the world's first $5 trillion company (late 2025), operating a near-monopoly on advanced AI chips.",
      "status": "hit",
      "weight": 0.5,
      "ordinal": -7,
      "source_id": "SEM_011",
      "expected_date": "2026-04-29",
      "observed_date": "2026-04-29"
    },
    {
      "kind": "prereq",
      "label": "Nvidia Data Center revenue +66% YoY, contributing ~90% of $57B fiscal Q3 revenue; >$4.5T market cap entirely underpinned by AI silicon.",
      "status": "hit",
      "weight": 0.5,
   
... (truncated)