Sam Altman predicted 40x year-over-year hyperdeflation of AI costs at constant capability.

Sam Altman predicted 40x year-over-year hyperdeflation of AI costs at constant capability. | This goes handinhand with what we've talked about in the past, Sam Alman's comment about 40x year-over-year hyperdelation of costs at constant capability.

This goes handinhand with what we've talked about in the past, Sam Alman's comment about 40x year-over-year hyperdelation of costs at constant capability.

Beneficiaries (23)

Adverse (6)

{
  "nia": false,
  "qty": "40x per year cost reduction",
  "url": "https://www.youtube.com/watch?v=T8X6kp-pcKs",
  "mode": "CITED_PREDICTION",
  "role": "Cited-Executive",
  "context": "the capability density of models is increasing. This goes handinhand with what we've talked about in the past, Sam Alman's comment about 40x year-over-year hyperdelation of costs at constant capability.",
  "to_year": 2027,
  "cited_by": "Alex Wissner-Gross",
  "verbatim": "This goes handinhand with what we've talked about in the past, Sam Alman's comment about 40x year-over-year hyperdelation of costs at constant capability.",
  "direction": "DOWN",
  "from_year": 2026,
  "timeframe": "annual, ongoing",
  "conv_level": "HIGH",
  "milestones": [
    {
      "kind": "prereq",
      "label": "Nvidia quadrupled chip production output while only doubling human headcount — achieved by deploying AI coding tools (Cursor, Claude Code) a",
      "status": "hit",
      "weight": 0.5,
      "ordinal": -8,
      "source_id": "SEM_012",
      "expected_date": "2026-04-29",
      "observed_date": "2026-04-29"
    },
    {
      "kind": "prereq",
      "label": "Training runs costing $10 billion for a single model will commence sometime in 2025.",
      "status": "hit",
      "weight": 0.5,
      "ordinal": -7,
      "source_id": "SEM_008",
      "expected_date": "2026-04-29",
      "observed_date": "2026-04-29"
    },
    {
      "kind": "prereq",
      "label": "2025 will be the definitive year that agentic systems finally hit the mainstream.",
      "status": "hit",
      "weight": 0.5,
      "ordinal": -6,
      "source_id": "SEM_042",
      "expected_date": "2026-04-29",
      "observed_date": "2026-04-29"
    },
    {
      "kind": "prereq",
      "label": "Recursive self-improvement is already happening now (no longer three years out)",
      "status": "hit",
      "weight": 0.5,
      "ordinal": -5,
      "source_id": "238_009",
      "expected_date": "2026-04-29",
      "observed_date": "2026-04-29"
    },
    {
      "kind": "llm_pre_event",
      "label": "OpenAI / Anthropic / Google headline frontier-model token prices fall 60-80% YoY",
      "notes": "HIT-direction - Industry pricing data shows ~80% YoY drop, far below Altman's 40x but consistent with capability-deflation directional claim. Headline GPT-5.2 input ~$1.75/MTok vs GPT-4 input ~$30/MTok in 2024.",
      "source": "https://tokenmix.ai/blog/ai-api-pricing-war-2026; https://intuitionlabs.ai/articles/ai-api-pricing-comparison-grok-gemini-openai-claude",
      "status": "hit",
      "weight": 0.4,
      "ordinal": -4,
      "source_id": null,
      "confidence": 0.9,
      "source_url": "https://tokenmix.ai/blog/ai-api-pricing-war-2026",
      "expected_date": "2026-04-30",
      "observed_date": "2026-04-30",
      "research_origin": "deep_research",
      "measurement_criterion": "Industry pricing trackers (TokenMix, IntuitionLabs, Silicon Data) confirm headline LLM API prices for frontier-class models down 60-80% year-over-year between 2025 and 2026"
    },
    {
      "kind": "llm_pre_event",
      "label": "Same-capability constant-quality token cost falls >=10x in a single year",
      "notes": "Constant-capability is the right Altman framing. 40x annual is the upper bar; 10x+ is the academic-consensus floor.",
      "source": "https://artificialanalysis.ai; https://www.silicondata.com/blog/llm-cost-per-token",
      "status": "pending",
      "weight": 0.4,
      "ordinal": -3,
      "source_id": null,
      "confidence": 0.85,
      "source_url": "https://www.silicondata.com/blog/llm-cost-per-token",
      "expected_date": "2026-12-15",
      "research_origin": "deep_research",
      "expected_date_range": {
        "to": "2027-06-30",
        "from": "2026-06-01"
      },
      "measurement_criterion": "Independent benchmark (e.g., Artificial Analysis, ARC-AGI cost-per-task) shows the cheapest model that matches GPT-4-Turbo (Mar-2024) on key benchmarks costs >=10x less per outpu
... (truncated)