← Cockpit
FUT_004predictionAILarge-Action-Models-displace-LLMs

Rapid ascension and enterprise dominance of Large Action Models (LAMs) by 2026-2031 — LLM as primary digital interface becomes obsolete; LAMs observe complex behavioral patterns, internalize multi-step business processes, predict physical-robot mainten...

Predictor: Amy Webb

Prior probability
75.0%
Current probability
61.6%
evolves via intake + LBP
Conviction
5/5
Signal quality
B
Resolution
in_progress
Window
2026-01-01 – 2031-09-30
Edges in / out
6 / 0
Tickers exposed
4

Prediction text

Rapid ascension and enterprise dominance of Large Action Models (LAMs) by 2026-2031 — LLM as primary digital interface becomes obsolete; LAMs observe complex behavioral patterns, internalize multi-step business processes, predict physical-robot maintenance, and independently execute strategic operational decisions without continuous human prompting. Obsolescence of prompt engineering; shift to managing fleets of autonomous software operators. | First Fortune 500 LAM-native operational deployment

Key catalyst: First Fortune 500 LAM-native operational deployment

Watch events: Enterprise LAM deployment announcements; prompt-engineering job market trends

Resolution evidence

Status: in_progress

Webb FTSG Tech Trends 2026 publication formalizes LAM framework; Rabbit R1, Humane Pin, Adept ACT-1, MultiOn validate action-model paradigm.

Predictor: Amy Webb

κ + Brier as of 2026-05-22
κ (discount)
0.500
Brier
Hits / Misses
0 / 0
Hit rate

Evidence about this node from Amy Webb 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 via embedding similarity 0.633

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

Base rate
20.0%
1/5 historical
Inside weight
Outside weight
no pull
inside 61.6% → blend 61.6% 0.0pp)

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 75%2026-04-302026-05-022026-05-24
intake v2milestone miss sweeplbp propagationreference class assignedlegacy v1prior_prob (analyst seed)current = 61.6%

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 ✓ · 1 overdue ⏱ · 8 pending
  1. 2025-09-01 → 2026-06-30overdueSalesforce xLAM family release (xGen-Sales + xLAM models 1B-176B)
    How: Salesforce publishes the xLAM model family (1B-8x22B parameter range) and xGen-Sales, with documentation of action-execution evaluation feedback loops, and lists at least one Fortune 500 customer using it operationally.
    Source: Salesforce — Large Action Models (xLAM)conf 85%
    Notes: Mainline enterprise vendor productization. Model release confirmed; the open question is named Fortune 500 deployment with operational decision authority.
  2. 2026-02-10hitGenesys announces LAM-powered agentic virtual agent (Feb 10, 2026)
    How: Genesys publicly announces production-grade Large Action Model (LAM)-powered agentic virtual agent product with general availability targeted for Q1 FY2027.
    Source: InformationWeek — 2026 enterprise AI predictions: fragmentation, commodification, and the agent push facing CIOsconf 92%
    Notes: HIT — first vendor publicly framing a product as LAM-class for enterprise, key threshold for the 2026 timeline.
  3. 2027-01-31pendingQ1 window check-in (25%)
  4. 2026-06-01 → 2027-12-31pendingFirst Fortune 500 disclosure of LAM-native operational deployment (no human-in-loop authority)
    How: A Fortune 500 company publicly states in 10-K, earnings call, or press release that an action-class AI model autonomously executes a multi-step business process (e.g., procurement, IT ticketing, fleet maintenance) without per-action human approval.
    Source: Anticipated — public 10-K filings, AI vendor case studies, enterprise CIO conferencesconf 50%
    Notes: Direct test of 'independently execute strategic operational decisions' clause. 2026-2027 window aligns with Genesys + Salesforce rollouts.
  5. 2026-09-01 → 2028-06-30pendingIndependent benchmark shows LAM > LLM on multi-step enterprise tasks
    How: A peer-reviewed or major industry benchmark (e.g., AgentBench, OSWorld, or successor) reports LAM-class systems materially outperforming LLM-only baselines on >10-step business workflows with statistical significance.
    Source: Anticipated — academic benchmarks, MLPerf, or AI safety institute evaluationsconf 55%
    Notes: Required to establish that LAMs functionally displace LLMs as the primary digital interface.
  6. 2028-03-02pendingQ2 window check-in (50%)
  7. 2027-01-01 → 2029-06-30pendingPublic market disclosure of prompt-engineering job decline / LAM-ops job creation
    How: BLS, LinkedIn Economic Graph, or Indeed Hiring Lab reports a measurable decline in 'prompt engineer' role postings alongside rise in 'AI agent operator' / 'autonomous workflow manager' roles.
    Source: Anticipated — labor market research from BLS, LinkedIn Economic Graphconf 45%
    Notes: Cascade — labor-market signal that 'shift to managing fleets of autonomous software operators' is real.
  8. 2029-04-02pendingQ3 window check-in (75%)

No downstream cascades — this prediction is a leaf in the dependency graph.

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: 62%)

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
LBP2026-05-24T02:00:02Z61.6%+2.3pp
Network propagation: 59.3% → 61.6%
4-iter LBP, residual 0.01000 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run 806b02f8
LBP2026-05-17T02:00:01Z59.3%+4.7pp
Network propagation: 54.7% → 59.3%
5-iter LBP, residual 0.00689 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run e607fa96
LBP2026-05-10T02:00:02Z54.7%+9.5pp
Network propagation: 45.1% → 54.7%
6-iter LBP, residual 0.00584 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run e5c18d29
LBP2026-05-03T02:00:01Z45.1%+17.2pp
Network propagation: 27.9% → 45.1%
6-iter LBP, residual 0.00677 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run 1a683ac9
metadata_milestone_miss_sweep2026-05-02T22:07:21Z27.9%-23.9pp
metadata_milestone_miss_sweep bayesian_v2 n=1 inside=0.475 blend=0.279 LLR=-0.172 κ=0.50 w_in=0.34 agi_breakthrough_5y
Raw metadata
{
  "trf": 0.9418866657386002,
  "kappa": 0.5,
  "base_rate": 0.2,
  "predictor": "Amy Webb",
  "total_llr": -0.4054651081081644,
  "grace_days": 7,
  "bayesian_v2": true,
  "prior_logit": 0.07262063684461216,
  "bayes_factor": "1.2:1 against",
  "blend_reason": "blend 34% inside / 65% outside (TRF=0.942, base_rate=0.200 from agi_breakthrough_5y)",
  "inside_prior": 0.5181471845920733,
  "kappa_source": "predictor_table",
  "n_milestones": 1,
  "blend_applied": true,
  "contributions": [
    {
      "llr": -0.4054651081081644,
      "kind": "llm_pre_event",
      "kappa": 0.425,
      "label": "Salesforce xLAM family release (xGen-Sales + xLAM models 1B-176B)",
      "weight": 0.4,
      "strength": "weak",
      "confidence": 0.85,
      "source_url": "https://www.salesforce.com/agentforce/large-action-models/",
      "adjusted_llr": -0.17232267094596987,
      "expected_date": "2026-01-30",
      "measurement_criterion": "Salesforce publishes the xLAM model family (1B-8x22B parameter range) and xGen-Sales, with documentation of action-execution evaluation feedback loops, and lists at least one Fortune 500 customer using it operationally."
    }
  ],
  "evidence_kind": "metadata_milestone_miss_sweep",
  "inside_source": "history_v2",
  "inside_weight": 0.3406793339829798,
  "outside_weight": 0.6593206660170202,
  "posterior_prob": 0.2792914542360655,
  "posterior_logit": -0.09970203410135771,
  "predictor_brier": null,
  "inside_posterior": 0.47509511862948256,
  "blended_posterior": 0.2792914542360655,
  "reference_class_id": "agi_breakthrough_5y",
  "total_adjusted_llr": -0.17232267094596987,
  "predictor_n_resolved": 0
}
LBP2026-04-30T16:39:51Z51.8%+15.1pp
Network propagation: 36.7% → 51.8%
5-iter LBP, residual 0.00825 · damping 0.5, w_intrinsic 0.5 · method lbp_v2 · run 0c8a4ea3
legacy v12026-04-30T16:13:50Z36.7%-15.1pp
reference_class_assigned bayesian_v2 inside=0.750 blend=0.367 w_in=0.34 agi_breakthrough_5y
LBP2026-04-30T02:18:57Z51.8%+15.1pp
Network propagation: 36.7% → 51.8%
5-iter LBP, residual 0.00825 · damping 0.5, w_intrinsic 0.5 · method lbp_v1 · run 592311ef
legacy v12026-04-30T01:56:50Z36.7%-38.3pp
reference_class_assigned bayesian_v2 inside=0.750 blend=0.367 w_in=0.34 agi_breakthrough_5y

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
killerTK06
China-Taiwan Military Conflict
8.0%0.0500.750+0.078
killerTK03
AI Regulatory Moratorium (EU/US Capability Freeze)
10.0%0.0500.750+0.064
killerTK11
Autonomous Regulatory Block (Level 4 Halt)
10.0%0.0500.750+0.064
killerTK01
AGI Capability Plateau (2026-27 Training Stall)
15.0%0.0500.750+0.029

Top outgoing (children)

Predictions THIS node influences

No outgoing edges.

Ticker exposure

4 ticker(s) linked

Adverse (4)

ALLPGRTRVUBER

Prerequisites (6)

Predictions that must hit first
TypePredTitleDomainLag
correlateS_HUMANOID_ENTERPRISE_2028Humanoid R2: 100K+ enterprise by Nov 2028humanoid_deployment
correlateS_AGI_FAST_2027AGI fast: drop-in remote worker by 2027-09agi_general_capability
killerTK01AGI Capability Plateau (2026-27 Training Stall)
killerTK03AI Regulatory Moratorium (EU/US Capability Freeze)
killerTK11Autonomous Regulatory Block (Level 4 Halt)
killerTK06China-Taiwan Military Conflict

Dependents (0)

Predictions enabled by this
TypePredTitleDomainLag
No dependents

Validations (1)

Resolution events
Observed atStatusByNotes
2026-04-29partialthesis_timeline_v1.0_importWebb FTSG Tech Trends 2026 publication formalizes LAM framework; Rabbit R1, Humane Pin, Adept ACT-1, MultiOn validate action-model paradigm.

Linked documents (10)

Auto-generated by cosine similarity from Polymarket / Manifold / EDGAR / GDELT

Raw metadata

From Thesis_Timeline_v1.0_FINAL workbook
{
  "nia": false,
  "mode": "FORECAST",
  "role": "Cited-Other",
  "context": "First Amy Webb entry. FTSG 2026 Tech Trends report specific LAM framing. Couples with AUT_017 (Ba agentic RL), AUT_021 (Andreessen data entropy), IND_026 (Weil scaffolding eaten), SPC_027 (Finn trillion-dollar agent economy).",
  "to_year": 2031,
  "conv_cues": "futurist FIRST_PERSON; coined framework LAM",
  "direction": "HAPPEN",
  "from_year": 2026,
  "timeframe": "2026-2031",
  "conv_level": "HIGH",
  "milestones": [
    {
      "kind": "llm_pre_event",
      "label": "Salesforce xLAM family release (xGen-Sales + xLAM models 1B-176B)",
      "notes": "Mainline enterprise vendor productization. Model release confirmed; the open question is named Fortune 500 deployment with operational decision authority.",
      "source": "Salesforce — Large Action Models (xLAM)",
      "status": "overdue",
      "weight": 0.4,
      "ordinal": -10,
      "source_id": null,
      "confidence": 0.85,
      "source_url": "https://www.salesforce.com/agentforce/large-action-models/",
      "expected_date": "2026-01-30",
      "miss_emitted_at": "2026-05-02T22:07:21.384228+00:00",
      "miss_emitted_by": "metadata_milestone_sweep",
      "research_origin": "deep_research",
      "expected_date_range": {
        "to": "2026-06-30",
        "from": "2025-09-01"
      },
      "measurement_criterion": "Salesforce publishes the xLAM model family (1B-8x22B parameter range) and xGen-Sales, with documentation of action-execution evaluation feedback loops, and lists at least one Fortune 500 customer using it operationally."
    },
    {
      "kind": "llm_pre_event",
      "label": "Genesys announces LAM-powered agentic virtual agent (Feb 10, 2026)",
      "notes": "HIT — first vendor publicly framing a product as LAM-class for enterprise, key threshold for the 2026 timeline.",
      "source": "InformationWeek — 2026 enterprise AI predictions: fragmentation, commodification, and the agent push facing CIOs",
      "status": "hit",
      "weight": 0.4,
      "ordinal": -9,
      "source_id": null,
      "confidence": 0.92,
      "source_url": "https://www.informationweek.com/machine-learning-ai/2026-enterprise-ai-predictions-fragmentation-commodification-and-the-agent-push-facing-cios",
      "expected_date": "2026-02-10",
      "observed_date": "2026-02-10",
      "research_origin": "deep_research",
      "measurement_criterion": "Genesys publicly announces production-grade Large Action Model (LAM)-powered agentic virtual agent product with general availability targeted for Q1 FY2027."
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q1 window check-in (25%)",
      "status": "pending",
      "weight": 0.05,
      "ordinal": -8,
      "source_id": null,
      "expected_date": "2027-01-31",
      "observed_date": null
    },
    {
      "kind": "llm_pre_event",
      "label": "First Fortune 500 disclosure of LAM-native operational deployment (no human-in-loop authority)",
      "notes": "Direct test of 'independently execute strategic operational decisions' clause. 2026-2027 window aligns with Genesys + Salesforce rollouts.",
      "source": "Anticipated — public 10-K filings, AI vendor case studies, enterprise CIO conferences",
      "status": "pending",
      "weight": 0.4,
      "ordinal": -7,
      "source_id": null,
      "confidence": 0.5,
      "expected_date": "2027-03-17",
      "research_origin": "training",
      "expected_date_range": {
        "to": "2027-12-31",
        "from": "2026-06-01"
      },
      "measurement_criterion": "A Fortune 500 company publicly states in 10-K, earnings call, or press release that an action-class AI model autonomously executes a multi-step business process (e.g., procurement, IT ticketing, fleet maintenance) without per-action human approval."
    },
    {
      "kind": "llm_pre_event",
      "label": "Independent benchmark shows LAM > LLM on multi-step enterprise tasks",
      "notes": "Required to establish that
... (truncated)