← Cockpit
CMQ_001predictionAIAGI-capability-roadmap

By 2026, AI will reach 'intern-level' capability — millions of virtual interns performing supervised, economically useful tasks.

Predictor: Sam Altman

Prior probability
72.0%
Current probability
44.8%
evolves via intake + LBP
Conviction
5/5
Signal quality
A
Resolution
pending
Window
2026-01-01 – 2026-09-30
Edges in / out
5 / 14
Tickers exposed
13

Prediction text

By 2026, AI will reach 'intern-level' capability — millions of virtual interns performing supervised, economically useful tasks. | OpenAI GPT-5/6-class agent releases

Key catalyst: OpenAI GPT-5/6-class agent releases

Watch events: OpenAI agent product releases; GDPval-style benchmarks; labor-substitution evidence from early enterprise deployments.

Resolution evidence

Status: pending

Consistent with OpenAI agent releases (Operator, Deep Research, ChatGPT agents) approaching intern-level task execution through 2025-2026.

Predictor: Sam Altman

κ + Brier as of 2026-05-22
κ (discount)
0.583
Brier
0.0625
excellent
Hits / Misses
0 / 0
of 1 resolved
Hit rate
0.0%
Calibration plot (stated vs observed)

Evidence about this node from Sam Altman 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: agi_breakthrough_5y

Linked

Major capability discontinuity (e.g. AGI by named target year, 5-year horizon)

Base rate
20.0%
1/5 historical
Inside weight
0.686
TRF=0.45
Outside weight
0.314
pulling toward base rate
inside 58.1% → blend 44.8% -13.4pp)

Tetlock-style outside view: at TRF=1 (just predicted), outside view dominates (w_in=0.3). At TRF=0 (deadline), inside view dominates (w_in=1.0). The blend regularizes overconfident inside views toward the historical base rate.

Probability over time

9 prob_history rows
0%25%50%75%100%prior 72%2026-04-302026-05-032026-05-30
intake v2milestone miss sweeplbp propagationreference class assignedlegacy v1prior_prob (analyst seed)current = 44.8%

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: 1 fired ✓ · 3 overdue ⏱
  1. 2026-02-14overdueQ1 window check-in (25%)
  2. 2026-03-30overdueQ2 window check-in (50%)
  3. 2026-05-13overdueQ3 window check-in (75%)

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: 45%)

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
metadata_milestone_miss_sweep2026-05-30T22:15:00Z44.8%-19.0pp
metadata_milestone_miss_sweep bayesian_v2 n=1 inside=0.581 blend=0.448 LLR=-0.237 κ=0.58 w_in=0.69 agi_breakthrough_5y
Raw metadata
{
  "trf": 0.4487974555581937,
  "kappa": 0.5833,
  "base_rate": 0.2,
  "predictor": "Sam Altman",
  "total_llr": -0.4054651081081644,
  "grace_days": 7,
  "bayesian_v2": true,
  "prior_logit": 0.5653088523306947,
  "bayes_factor": "1.3:1 against",
  "blend_reason": "blend 68% inside / 31% outside (TRF=0.449, base_rate=0.200 from agi_breakthrough_5y)",
  "inside_prior": 0.6376800141997431,
  "kappa_source": "predictor_table",
  "n_milestones": 1,
  "blend_applied": true,
  "contributions": [
    {
      "llr": -0.4054651081081644,
      "kind": "quartile_checkpoint",
      "kappa": 0.5833,
      "label": "Q3 window check-in (75%)",
      "weight": 0.05,
      "strength": "weak",
      "confidence": null,
      "source_url": null,
      "adjusted_llr": -0.2365077975594923,
      "expected_date": "2026-05-13",
      "measurement_criterion": null
    }
  ],
  "evidence_kind": "metadata_milestone_miss_sweep",
  "inside_source": "history_v2",
  "inside_weight": 0.6858417811092643,
  "outside_weight": 0.3141582188907357,
  "posterior_prob": 0.4476895519386729,
  "posterior_logit": 0.3288010547712024,
  "predictor_brier": 0.0625,
  "inside_posterior": 0.5814676264348703,
  "blended_posterior": 0.4476895519386729,
  "reference_class_id": "agi_breakthrough_5y",
  "total_adjusted_llr": -0.2365077975594923,
  "predictor_n_resolved": 1
}
LBP2026-05-24T02:00:02Z63.8%-5.1pp
Network propagation: 68.9% → 63.8%
4-iter LBP, residual 0.01000 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run 806b02f8
intake_event_update2026-05-21T23:15:16Z68.9%+9.2pp
intake:7afeeb9a-f217-4dd2-b910-24ff14bdfc39 bayesian_v2 inside=0.689 blend=0.689 LLR=0.404 κ=0.58 no_blend
Raw metadata
{
  "trf": 0.481731855298543,
  "kappa": 0.5833,
  "base_rate": null,
  "predictor": "Sam Altman",
  "total_llr": 0.6931471805599453,
  "bayesian_v2": true,
  "prior_logit": 0.3904375358084699,
  "bayes_factor": "1.5:1 favoring",
  "blend_reason": "no reference_class linked",
  "inside_prior": 0.5963880226965899,
  "kappa_source": "predictor_table",
  "blend_applied": false,
  "contributions": [
    {
      "llr": 0.6931471805599453,
      "kappa": 0.5833,
      "label": "Weeks-long autonomous task capability is well beyond 'intern-level'.",
      "adjusted_llr": 0.4043127504206161
    }
  ],
  "evidence_kind": "intake_event_update",
  "inside_source": "history_v2",
  "inside_weight": 1,
  "outside_weight": 0,
  "posterior_prob": 0.6888503979684238,
  "evidence_origin": "daily_intake",
  "llm_suggestions": [
    {
      "polarity": "corroborates",
      "status_change": "unchanged",
      "evidence_strength": "moderate",
      "delta_prob_suggestion": 0.05
    }
  ],
  "posterior_logit": 0.7947502862290861,
  "predictor_brier": 0.0625,
  "evidence_doc_ids": [],
  "inside_posterior": 0.6888503979684238,
  "blended_posterior": 0.6888503979684238,
  "reference_class_id": null,
  "total_adjusted_llr": 0.4043127504206161,
  "predictor_n_resolved": 1
}
LBP2026-05-10T02:00:02Z59.6%-1.7pp
Network propagation: 61.4% → 59.6%
6-iter LBP, residual 0.00584 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run e5c18d29
LBP2026-05-03T02:00:01Z61.4%-3.1pp
Network propagation: 64.4% → 61.4%
6-iter LBP, residual 0.00677 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run 1a683ac9
metadata_milestone_miss_sweep2026-05-02T22:07:21Z64.4%-10.0pp
metadata_milestone_miss_sweep bayesian_v2 n=2 inside=0.644 blend=0.644 LLR=-0.473 κ=0.58 no_blend
Raw metadata
{
  "trf": 0.5517581791161151,
  "kappa": 0.5833,
  "base_rate": null,
  "predictor": "Sam Altman",
  "total_llr": -0.8109302162163288,
  "grace_days": 7,
  "bayesian_v2": true,
  "prior_logit": 1.0675603953538413,
  "bayes_factor": "1.6:1 against",
  "blend_reason": "no reference_class linked",
  "inside_prior": 0.7441326937015715,
  "kappa_source": "predictor_table",
  "n_milestones": 2,
  "blend_applied": false,
  "contributions": [
    {
      "llr": -0.4054651081081644,
      "kind": "quartile_checkpoint",
      "kappa": 0.5833,
      "label": "Q1 window check-in (25%)",
      "weight": 0.05,
      "strength": "weak",
      "confidence": null,
      "source_url": null,
      "adjusted_llr": -0.2365077975594923,
      "expected_date": "2026-02-14",
      "measurement_criterion": null
    },
    {
      "llr": -0.4054651081081644,
      "kind": "quartile_checkpoint",
      "kappa": 0.5833,
      "label": "Q2 window check-in (50%)",
      "weight": 0.05,
      "strength": "weak",
      "confidence": null,
      "source_url": null,
      "adjusted_llr": -0.2365077975594923,
      "expected_date": "2026-03-30",
      "measurement_criterion": null
    }
  ],
  "evidence_kind": "metadata_milestone_miss_sweep",
  "inside_source": "history_v2",
  "inside_weight": 0.6137692746187193,
  "outside_weight": 0.38623072538128067,
  "posterior_prob": 0.6444072531106255,
  "posterior_logit": 0.5945448002348567,
  "predictor_brier": 0.0625,
  "inside_posterior": 0.6444072531106255,
  "blended_posterior": 0.6444072531106255,
  "reference_class_id": null,
  "total_adjusted_llr": -0.4730155951189846,
  "predictor_n_resolved": 1
}
legacy v12026-04-30T19:17:54Z74.4%+8.4pp
intake:99aa73db-75b1-4b1e-8470-a11f87b23937 bayesian_v2 inside=0.744 blend=0.744 LLR=0.404 κ=0.58 no_blend
LBP2026-04-30T16:39:51Z66.0%-2.1pp
Network propagation: 68.1% → 66.0%
5-iter LBP, residual 0.00825 · damping 0.5, w_intrinsic 0.5 · method lbp_v2 · run 0c8a4ea3
LBP2026-04-30T02:18:57Z68.1%-3.9pp
Network propagation: 72.0% → 68.1%
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
killerTK03
AI Regulatory Moratorium (EU/US Capability Freeze)
10.0%0.0500.720+0.205
killerTK01
AGI Capability Plateau (2026-27 Training Stall)
15.0%0.0500.720+0.172
prereq238_009
Recursive self-improvement is already happening now (no longAlex Wissner-Gross
78.1%0.7200.050+0.120

Top outgoing (children)

Predictions THIS node influences

KindNodeTheir probP(c|s=T)P(c|s=F)Δ implied
prereq241_057
Elon Musk believes robot building robot is imminentElon Musk
44.8%0.5500.050-0.143
prereq242_001
Elon's Terafab will build 1 terawatt of AI compute per year,Elon Musk
43.9%0.5500.050-0.134
prereq233_021
AI learning will improve via closed-loop reinforcement learnJoe Liemandt
38.7%0.4500.050-0.133
prereq237_023
Baby AGI agents will need and develop an 'immune system' forAlex Wissner-Gross
40.7%0.5000.050-0.128
prereq238_052
$100 trillion companies within 5 years (3 years from now, peElon Musk
41.7%0.5500.050-0.112

Ticker exposure

13 ticker(s) linked

Beneficiaries (13)

BBAINVDAGTLBSOUNAIMETAMSFTORCLTCEHYAMZNBABAGOOGLIBM

Prerequisites (5)

Predictions that must hit first
TypePredTitleDomainLag
prereq238_009Recursive self-improvement is already happening now (no longer three years out)AI
correlateS_AGI_MID_2029AGI mid: Kurzweil 2029 pathagi_general_capability
correlateS_AI_PAUSE_2026Major-country AI pause beginning 2026ai_regulatory_pause
killerTK01AGI Capability Plateau (2026-27 Training Stall)
killerTK03AI Regulatory Moratorium (EU/US Capability Freeze)

Dependents (14)

Predictions enabled by this
TypePredTitleDomainLag
prereq235_030Ray Kurzweil predicts Longevity Escape Velocity (LEV) by 2033.Biotech/Longevity
prereq241_043ASI will arrive within 2 years to 5 years to this next decadeAI
prereqCMQ_002By 2028, AI systems will reach 'independent researcher' level — driving autonomous scientific discoveries without human intervention.AI
prereq232_047Mass drivers on the moon will shoot AI satellites into deep space; self-sustaining lunar city will follow.Space
prereq239_009People will be on Mars within 10 yearsSpace
prereq241_057Elon Musk believes robot building robot is imminentRobotics
prereq242_001Elon's Terafab will build 1 terawatt of AI compute per year, 50x current global productionAI
prereqSEM_034True artificial general intelligence will be achieved between 2032 and 2042 — 'first we solve AI, then use AI to solve everything else'.AI/AGI
prereq238_052$100 trillion companies within 5 years (3 years from now, per Diamandis interpretation of Musk)Markets/Stocks
prereq239_008Moon base will exist in 10 yearsSpace
prereq237_023Baby AGI agents will need and develop an 'immune system' for prompt injection and cybersecurity threats in real time.AI
prereq239_010Mass driver on the moon within 10 yearsSpace
prereq233_021AI learning will improve via closed-loop reinforcement learning cycle making results keep increasing.AI
prereq230_022Elon plans to produce tens of millions of robots per year in just a few years.Robotics

Linked documents (10)

Auto-generated by cosine similarity from Polymarket / Manifold / EDGAR / GDELT
SimSourceTitleMarket probPolarityReviewedPublished
0.750codex_research_packMETR - Measuring AI Ability to Complete Long Taskscorroboratespending2025-03-19
0.750codex_research_packOECD - Exploring Possible AI Trajectories Through 2030corroboratespending2026-04-26
0.720manifoldWill OpenAI release a GPT version > 5.5 before June 2026?11%mentionspending2026-05-15
0.685manifoldat the end of 2026, will an AI be able to generate a full high-quality tv ep to a prompt?35%mentionspending2026-05-31
0.650manifoldWill "How Go Players Disempower Themselves to AI" make the top fifty posts in LessWrong's 2026 Annual Review?34%mentionspending2026-05-02
0.635manifoldWill OpenAI de-deploy GPT-5.5 before 2027 for safety, security, cyber-risk, or other threat-related reasons?8%mentionspending2026-05-04
0.624github_releaseopenai/openai-python v2.23.0mentionspending2026-02-24
0.611manifoldGPT 5.6 released by…?mentionspending2026-05-18
0.606manifoldWhen will Indonesia announce the IMO 2026 Team?mentionspending2026-04-27
0.603manifoldWhat will I get at IMO 2026?mentionspending2026-05-03

Raw metadata

From Thesis_Timeline_v1.0_FINAL workbook
{
  "nia": false,
  "qty": "intern-level",
  "mode": "FORECAST",
  "role": "Cited-CEO",
  "context": "Altman's capability roadmap: 2026 intern-level; 2028 independent researcher; 2030 surpass peak human expert in all cognitive domains. Framed AGI as a 'phase shift' rather than a finish line.",
  "to_year": 2026,
  "conv_cues": "will reach; specific year; CEO FIRST_PERSON",
  "direction": "HAPPEN",
  "from_year": 2026,
  "timeframe": "by 2026",
  "conv_level": "HIGH",
  "milestones": [
    {
      "kind": "quartile_checkpoint",
      "label": "Q1 window check-in (25%)",
      "status": "overdue",
      "weight": 0.05,
      "ordinal": -4,
      "source_id": null,
      "expected_date": "2026-02-14",
      "observed_date": null,
      "miss_emitted_at": "2026-05-02T22:07:21.384228+00:00",
      "miss_emitted_by": "metadata_milestone_sweep"
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q2 window check-in (50%)",
      "status": "overdue",
      "weight": 0.05,
      "ordinal": -3,
      "source_id": null,
      "expected_date": "2026-03-30",
      "observed_date": null,
      "miss_emitted_at": "2026-05-02T22:07:21.384228+00:00",
      "miss_emitted_by": "metadata_milestone_sweep"
    },
    {
      "kind": "prereq",
      "label": "Recursive self-improvement is already happening now (no longer three years out)",
      "status": "hit",
      "weight": 0.5,
      "ordinal": -2,
      "source_id": "238_009",
      "expected_date": "2026-04-29",
      "observed_date": "2026-04-29"
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q3 window check-in (75%)",
      "status": "overdue",
      "weight": 0.05,
      "ordinal": -1,
      "source_id": null,
      "expected_date": "2026-05-13",
      "observed_date": null,
      "miss_emitted_at": "2026-05-30T22:15:00.756418+00:00",
      "miss_emitted_by": "metadata_milestone_sweep"
    },
    {
      "kind": "event",
      "label": "By 2026, AI will reach 'intern-level' capability — millions of virtual interns performing supervised, economically useful tasks.",
      "status": "pending",
      "weight": 1,
      "ordinal": 0,
      "source_id": "CMQ_001",
      "expected_date": "2026-06-26",
      "observed_date": null
    },
    {
      "kind": "cascade",
      "label": "Baby AGI agents will need and develop an 'immune system' for prompt injection and cybersecurity threats in real time.",
      "status": "pending",
      "weight": 0.5,
      "ordinal": 1,
      "source_id": "237_023",
      "expected_date": "2027-06-18",
      "observed_date": null
    },
    {
      "kind": "cascade",
      "label": "AI learning will improve via closed-loop reinforcement learning cycle making results keep increasing.",
      "status": "pending",
      "weight": 0.5,
      "ordinal": 2,
      "source_id": "233_021",
      "expected_date": "2028-06-22",
      "observed_date": null
    },
    {
      "kind": "cascade",
      "label": "Elon plans to produce tens of millions of robots per year in just a few years.",
      "status": "pending",
      "weight": 0.5,
      "ordinal": 3,
      "source_id": "230_022",
      "expected_date": "2029-12-10",
      "observed_date": null
    },
    {
      "kind": "cascade",
      "label": "$100 trillion companies within 5 years (3 years from now, per Diamandis interpretation of Musk)",
      "status": "pending",
      "weight": 0.5,
      "ordinal": 4,
      "source_id": "238_052",
      "expected_date": "2030-09-15",
      "observed_date": null
    },
    {
      "kind": "cascade",
      "label": "Ray Kurzweil predicts Longevity Escape Velocity (LEV) by 2033.",
      "status": "pending",
      "weight": 0.5,
      "ordinal": 5,
      "source_id": "235_030",
      "expected_date": "2033-07-30",
      "observed_date": null
    },
    {
      "kind": "cascade",
      "label": "Mass drivers on the moon will shoot AI satellites into deep space; self-sustaining lunar city will follow.",
      "status": "pending",
      "weight": 0.5,
     
... (truncated)