What this section answers
Posted rates are dated, the index is a composite rather than a workload predictor, the subsidy analysis is structural on the user side, and the forecast bets are judgments over a short record.
boundary · section 10 of 10
Evidence boundaries
What does this evidence not establish?
Boundary
This section is the reopen condition for every other part on the site. Read it before quoting a figure.
Source coordinate
Report heading Evidence boundaries, under Evidence boundaries, in The real cost of AI: August 2026.
The section itself
- All prices are evidence of posted rates on 2026-08-20, not durable claims. Provider pricing changes weekly.
- The quality-adjusted cost analysis uses the Artificial Analysis Intelligence Index, which is a composite benchmark score — it does not predict performance on a specific workload. The index is also an editorially-weighted nine-benchmark composite (text/English-centric, agentic-shifted since v4.1); version overhauls have moved scores by ~23 points, so cost-per-index-point figures are order-of-magnitude evidence, not precise rankings.
- The GPU pricing data is compiled from independent sources including two grade-A direct captures (getdeploying.com, spendark.com) and a compiled evidence file; both GPU claims are verified.
- The subsidy analysis is structural (the subsidy exists, it distorts decisions) and now has a cost-side quantitative anchor: frontier-lab list pricing runs below sustainable infrastructure return (OpenAI gross margin 33% in 2025 vs its own 46% forecast, inference costs ~$8.4B and 4x YoY; Anthropic 40%) — posted API prices remain effectively investor-subsidized. The user-side subsidized fraction is still unmeasured.
- The forecast bets above are dated, binary, probability-bearing judgments over a short record — not calibrated forecasts. FC4 and FC6 have named missing baselines; their probabilities are structured priors.
- Recheck by: 2026-11-20, or after a material model release, pricing change, or hardware generation shift.
Who uses this section, and for what
Developers and ML engineers
What the index-dependence caveat does to the quality-adjusted figure you are about to quote.
Engineering leaders and CTOs
Which of the six bets rest on structured priors rather than measured baselines.
Platform and infrastructure teams
How far the crossover bands travel with the assumptions behind them.
Researchers and data scientists
What the fine-tuning evidence does not settle.
Procurement and finance teams
Exactly which half of the subsidy claim is anchored and which is structural.
Executives, investors, and broad readers
The honest limit of a forecast made over a short record.
Reader questions that route through here
- How much of my AI spend is subsidized?Answered structurally, anchored on the cost side
- How do all these dimensions interact?Partially answered
In sequence