Progress in AI will continue as series of overlapping S-curves

Progress in AI will continue as series of overlapping S-curves | progress in AI is just a sort of series of you know sort of overlapping S-curves

progress in AI is just a sort of series of you know sort of overlapping S-curves

Beneficiaries (14)

Adverse (7)

{
  "nia": false,
  "url": "https://www.youtube.com/watch?v=N5KCm_55xeQ",
  "mode": "THESIS",
  "role": "Guest-CEO",
  "context": "progress in AI is just a sort of series of you know sort of overlapping S-curves",
  "verbatim": "progress in AI is just a sort of series of you know sort of overlapping S-curves",
  "direction": "UP",
  "timeframe": "ongoing",
  "conv_level": "HIGH",
  "milestones": [
    {
      "kind": "llm_pre_event",
      "label": "GPT-5.x family launched signalling completion of one S-curve",
      "notes": "HIT - GPT-5, 5.2, 5.3-Codex public; reasoning models (o-series) and Codex denote second S-curve onset.",
      "source": "https://i10x.ai/news/openai-2026-ai-roadmap-gpt-5-models; OpenAI blog",
      "status": "hit",
      "weight": 0.4,
      "ordinal": -7,
      "source_id": null,
      "confidence": 0.95,
      "source_url": "https://i10x.ai/news/openai-2026-ai-roadmap-gpt-5-models",
      "expected_date": "2026-04-30",
      "observed_date": "2026-02-25",
      "research_origin": "deep_research",
      "measurement_criterion": "OpenAI launches GPT-5 family (5, 5.2, 5.3-Codex) demonstrating the full pretrain-scaling S-curve completion before transitioning to reasoning-focused S-curve"
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q1 window check-in (25%)",
      "status": "pending",
      "weight": 0.05,
      "ordinal": -6,
      "source_id": null,
      "expected_date": "2026-07-28",
      "observed_date": null
    },
    {
      "kind": "llm_pre_event",
      "label": "Published AI capability progress chart shows visible second S-curve onset (reasoning models)",
      "source": "https://epoch.ai/data; Artificial Analysis quarterly reports",
      "status": "pending",
      "weight": 0.4,
      "ordinal": -5,
      "source_id": null,
      "confidence": 0.8,
      "source_url": "https://epoch.ai/data",
      "expected_date": "2026-09-15",
      "research_origin": "training",
      "expected_date_range": {
        "to": "2026-12-31",
        "from": "2026-06-01"
      },
      "measurement_criterion": "Independent benchmark trackers (Epoch AI, Artificial Analysis) publish visualizations showing two distinct S-curves: pretrain scaling plateau and reasoning/RL scaling onset"
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q2 window check-in (50%)",
      "status": "pending",
      "weight": 0.05,
      "ordinal": -4,
      "source_id": null,
      "expected_date": "2026-10-25",
      "observed_date": null
    },
    {
      "kind": "llm_pre_event",
      "label": "RL-from-verifiable-reward / reasoning S-curve produces measurable benchmark step-up",
      "notes": "Reasoning is the second S-curve Musk implicitly described.",
      "source": "FrontierMath, ARC-AGI, SWE-bench Verified leaderboards",
      "status": "pending",
      "weight": 0.4,
      "ordinal": -3,
      "source_id": null,
      "confidence": 0.85,
      "expected_date": "2026-12-15",
      "research_origin": "training",
      "expected_date_range": {
        "to": "2027-06-30",
        "from": "2026-06-01"
      },
      "measurement_criterion": "A reasoning-trained model achieves a >=15-pp benchmark gain (FrontierMath, ARC-AGI, SWE-bench Verified) over the prior pretrain-only state-of-the-art within a single model generation"
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q3 window check-in (75%)",
      "status": "pending",
      "weight": 0.05,
      "ordinal": -2,
      "source_id": null,
      "expected_date": "2027-01-22",
      "observed_date": null
    },
    {
      "kind": "llm_post_event",
      "label": "Public capability-progress narrative shifts from 'chinchilla scaling' to 'overlapping curves'",
      "source": "Mainstream press analysis; AI thought-leader X posts",
      "status": "pending",
      "weight": 0.4,
      "ordinal": -1,
      "source_id": null,
      "confidence": 0.55,
      "expected_date": "2027-01-30",
      "research_origin": "training",
      "e
... (truncated)