Data Hygiene

Surface data quality issues that would silently degrade inference. Each section is actionable — fix the root cause, then re-run the relevant downstream cron (LBP / CF refresh / link_docs_to_nodes).

Predictions w/ NULL prob/window
24
Out-of-range probs
0
Edges w/ NULL conditionals
0
Resolved-early w/o resolved_at
5
85 resolved-early total
Orphan ticker references
0
Isolated high-conv claims
28
conviction ≥4, no tickers, no scenario edges
Sparse scenarios (<8 linked)
3
lanes with <8 dots are hard to read
CF cache scenarios
37
latest 2026-05-24T04:31
Milestones w/o criterion
10196
6472/16668 have criterion · 1327 preds enriched
Ticker moves unmodeled
32999
0/32999 public impact links modeled · 0 staged
Backward prereqs
95
165 DROP correlates hidden · 22 negative correlates
Terminal w/o history
307
190 overdue active · 40 staged recalibrations
Source hosts unseeded
4
202/206 document hosts seeded

Predictions with NULL window or current_prob

Fix at intake or via /api/intake. Without these, the prediction is filtered out of every fork view.
  • 236_010 — 5% wealth tax would result in zero billionaires by next day[NULL window_end ]
  • 238_031 — Future is unpredictable beyond three weeks in the AI era[NULL window_end ]
  • 238_050 — Sci-fi tropes (warp drive, teleportation, time travel) are not yet achieved; rep[NULL window_end ]
  • 238_057 — Laws of economics and thermodynamics will still apply despite AI abundance[NULL window_end ]
  • 238_069 — The path to the abundant end-state will include diverse mutually-exclusive natio[NULL window_end ]
  • 239_017 — Future is 80%+ likely to be great[NULL window_end ]
  • 239_026 — Future will be very entertaining[NULL window_end ]
  • 240_023 — 45% of dementia cases are entirely preventable[NULL window_end ]
  • 240_052 — Uploading consciousness (whole brain emulation) really will be you in the future[NULL window_end ]
  • 243_046 — When Uber opens in a market, drunk driving incidents go down significantly[NULL window_end ]
  • 244_008 — If you build liquid supply with product-market fit, demand will show up[NULL window_end ]
  • 244_014 — AV is an industry where Chinese players have a lot to offer[NULL window_end ]
  • 244_021 — AI press will continue to dramatize machines replacing humans[NULL window_end ]
  • 244_030 — Physical world tech adoption always takes longer than digital world[NULL window_end ]
  • 244_034 — Companies get more conservative as they grow, but should do opposite[NULL window_end ]
  • 245_004 — EY estimates the extinct-species educational/consumer market at $1.7 trillion an[NULL window_end ]
  • 245_023 — Colossal considers de-extinction and species preservation a $10 trillion opportu[NULL window_end ]
  • 245_031 — 99% of synthetic biology focus will remain on human healthcare; similar tech app[NULL window_end ]
  • 245_032 — Anti-GMO perception of reintroduced Tasmanian tiger requires government educatio[NULL window_end ]
  • 245_034 — Wolves will not attack modern cattle due to current cattle-raising methods[NULL window_end ]

3 scenarios with <8 linked predictions

Fork lanes look thin. Run scripts/ops/backfill_scenario_prediction_edges.py with a lower --threshold (e.g. 0.58) to add more correlate edges.
  • S_IPO_TRILLION_NONE_5Y ($1T+ IPO) — 1 linked · No $1T+ IPO through 2031
  • S_ROBOTAXI_DELAYED (Robotaxi) — 4 linked · Robotaxi mass deployment delayed past 2031
  • S_COMPUTE_STARGATE_FAILURE (Compute scale) — 7 linked · Stargate failure: <500MW by 2029

28 high-conviction (≥4) claims with no ticker + no scenario edges

These are 'orphan' high-conviction claims. They don't appear in any tradeable fork view. Either link to a ticker or to a scenario family.
  • AUT_006 conv 5/5 — 2027 opening of the world's largest genetic BioVault in Dubai — infras
  • FUT_014 conv 5/5 — 'US political revolution' ranked #1 tier global risk 2026+ per Eurasia
  • IND_014 conv 5/5 — Generative AI has rendered static one-size-fits-all learning structura
  • INF_031 conv 5/5 — The biggest near-term AI risk to humanity is AI companies themselves a
  • IND_003 conv 5/5 — Latency of medical truth actively collapsing — genomic sequencing for
  • ROB_010 conv 5/5 — 'Watts are wealth' — as machines progressively take over physical prod
  • CYB_023 conv 5/5 — Immutable cryptographic verification layers — specifically Bitcoin — m
  • IND_026 conv 5/5 — Non-coders and engineers alike must build for 'where the models are go
  • CYB_010 conv 4/5 — Agentic commerce will radically redistribute — not grow — the global c
  • INF_030 conv 4/5 — By 2026, generative AI will democratize analytical and operational cap
  • CYB_006 conv 4/5 — When AI agents possess the ability to read, write, and restructure the
  • AI_029 conv 4/5 — Blockchain provenance protocols will be increasingly applied to enterp
  • AI_025 conv 4/5 — Most current generative AI wrappers are transient — they will fade int
  • CYB_014 conv 4/5 — The Exponential Organization (ExO) framework becomes mandatory corpora
  • CMQ_058 conv 4/5 — Localized hardware setups (multiple Apple Mac Studios, dedicated 'AI M
  • SPC_030 conv 4/5 — Eureka Labs vision: 'teacher + AI symbiosis' that scales human experti
  • AI_024 conv 4/5 — As AI-generated video becomes indistinguishable from reality, a massiv
  • AI_034 conv 4/5 — The AI-native education model will see global adoption within approxim
  • CYB_002 conv 4/5 — Despite raw agentic tool-use benchmarks tracking ahead of schedule, th
  • ROB_024 conv 4/5 — Automated robotics and AI-powered monitoring will establish a global n

Ticker impact links without approved move magnitudes

Run scripts/ops/stage_ticker_expected_moves.py to stage candidates. Promotion requires explicit metadata approval; existing links are preserved.

0 modeled · 32999 unknown · 0 staged candidates · 0 approved candidates.

Temporal prereq inversions (showing 20)

These edges may be valid evidence links, but as prereqs they imply a later event gates an earlier event. Review for reversal, reclassification, or conversion to a milestone.
  • 240_049 (2040-12-31) → 232_057 (2031-06-30)Last jobs to be automated will be government jobs, u / First person to arrive on moon/Mars will find bed ma
  • 234_008 (2040-12-31) → 232_057 (2031-06-30)Workflows will move from human-centric to agentic wi / First person to arrive on moon/Mars will find bed ma
  • 240_026 (2040-05-31) → 232_057 (2031-06-30)Elon's combined empire (SpaceX/xAI/Tesla/Optimus/Ter / First person to arrive on moon/Mars will find bed ma
  • 234_024 (2040-12-31) → 235_030 (2033-09-30)Financial systems between AI and blockchain will bec / Ray Kurzweil predicts Longevity Escape Velocity (LEV
  • 238_060 (2040-12-31) → 235_030 (2033-09-30)Universal basic AI (UBI for AI/compute) will be requ / Ray Kurzweil predicts Longevity Escape Velocity (LEV
  • 239_018 (2040-12-31) → 235_030 (2033-09-30)Universal High Income will be implemented / Ray Kurzweil predicts Longevity Escape Velocity (LEV
  • 245_038 (2040-12-31) → 235_030 (2033-09-30)People buying clones of their dogs is a profitable o / Ray Kurzweil predicts Longevity Escape Velocity (LEV
  • 240_030 (2040-11-30) → 235_030 (2033-09-30)Tesla's independent chip production would derisk pot / Ray Kurzweil predicts Longevity Escape Velocity (LEV
  • 230_050 (2038-06-30) → 232_057 (2031-06-30)One of the hidden Nvidia-style AI beneficiaries will / First person to arrive on moon/Mars will find bed ma
  • 236_044 (2040-09-30) → 235_030 (2033-09-30)Universities turning into incubators/venture studios / Ray Kurzweil predicts Longevity Escape Velocity (LEV
  • 236_038 (2040-09-30) → 235_030 (2033-09-30)Universities (Harvard, MIT, Stanford) will transitio / Ray Kurzweil predicts Longevity Escape Velocity (LEV
  • 236_043 (2040-05-31) → 235_030 (2033-09-30)Universities will become largest incubators on the p / Ray Kurzweil predicts Longevity Escape Velocity (LEV
  • 246_016 (2034-11-30) → COD_SPC_005 (2028-07-31)Dragonfly nuclear-powered octicopter arrives at Tita / NASA Dragonfly launches on Falcon Heavy in the July
  • 245_038 (2040-12-31) → 246_016 (2034-11-30)People buying clones of their dogs is a profitable o / Dragonfly nuclear-powered octicopter arrives at Tita
  • 234_008 (2040-12-31) → 246_016 (2034-11-30)Workflows will move from human-centric to agentic wi / Dragonfly nuclear-powered octicopter arrives at Tita
  • 240_049 (2040-12-31) → 246_016 (2034-11-30)Last jobs to be automated will be government jobs, u / Dragonfly nuclear-powered octicopter arrives at Tita
  • 236_044 (2040-09-30) → 246_016 (2034-11-30)Universities turning into incubators/venture studios / Dragonfly nuclear-powered octicopter arrives at Tita
  • 236_038 (2040-09-30) → 246_016 (2034-11-30)Universities (Harvard, MIT, Stanford) will transitio / Dragonfly nuclear-powered octicopter arrives at Tita
  • 236_043 (2040-05-31) → 246_016 (2034-11-30)Universities will become largest incubators on the p / Dragonfly nuclear-powered octicopter arrives at Tita
  • 240_026 (2040-05-31) → 246_016 (2034-11-30)Elon's combined empire (SpaceX/xAI/Tesla/Optimus/Ter / Dragonfly nuclear-powered octicopter arrives at Tita

Overdue active predictions needing adjudication or recalibration (showing 30)

These are not killed automatically. Use scripts/ops/stage_recalibration_candidates.py to stage timing-slip candidates for analyst review.
  • 240_005 conv 5 · ended 2026-06-30 · pendingThe organizational singularity will impact every company, nonprofit, and governm
  • 248_024 conv 5 · ended 2026-06-30 · pendingAI will drive an explosion of new religions and AI cults.
  • 231_026 conv 5 · ended 2026-06-30 · pendingSomething bad will definitely happen with OpenClaw agents by statistical chance.
  • 232_018 conv 5 · ended 2026-06-30 · pendingWe are living through a moment that will be in history for thousands of years: n
  • 248_050 conv 5 · ended 2026-06-30 · pendingOpus 4.7 removes manual dials; prompts become the new hyperparameters.
  • 241_018 conv 5 · ended 2026-06-30 · in_progressAlgorithms becoming more efficient will not reduce power demand - it will increa
  • 240_043 conv 4 · ended 2026-06-30 · pendingCS placements collapsed from 89% to 19% over 3 years; similar collapse coming fo
  • 236_046 conv 4 · ended 2026-06-30 · pendingSocial unrest unfortunately much closer than we'd like to think
  • 241_025 conv 4 · ended 2026-06-30 · pendingElon Musk predicts launch per hour cadence to populate satellite constellations
  • 236_031 conv 4 · ended 2026-06-30 · pendingWealth concentration will pick up as firms have fewer workers
  • 248_037 conv 4 · ended 2026-06-30 · pendingState data-center moratoriums will accelerate Dyson-swarm / orbital compute inno
  • 241_020 conv 4 · ended 2026-06-30 · pending1 GW of power = ~$50 billion in hardware/software/data centers
  • 231_025 conv 4 · ended 2026-06-30 · pendingEvery frontier lab will launch 24/7 agent offerings (table stakes).
  • 241_060 conv 4 · ended 2026-06-30 · pendingData centers will be built in space / rockets will carry data centers
  • 247_011 conv 4 · ended 2026-06-30 · pendingOpenAI user count will soon reach 1 billion
  • 241_056 conv 4 · ended 2026-06-30 · pendingElon Musk-style vertical integration is necessary to drive robotics costs down
  • 231_016 conv 4 · ended 2026-06-30 · pendingPermissionless agent activity is imminent; Claude-style approval prompts will go
  • 241_019 conv 4 · ended 2026-06-30 · pendingAI scaling laws have not hit an asymptote yet
  • 237_011 conv 4 · ended 2026-06-30 · pendingAI agents will have voices in the near future.
  • 230_026 conv 4 · ended 2026-06-30 · pendingThe QWERTY keyboard will persist until the heat death of the universe.

Resolved-early hits without metadata.resolved_at

59 hits known to be resolved-early; metadata.resolved_at backfill from prob_history pending.

5 need backfill. See Session 25 master-doc note on the 59-hit resolved-early case.