What this section answers
Six dated binary bets with explicit probabilities and annulment conditions — efficiency-driven decline, custom-silicon share, demand elasticity, subsidy contraction, open-weight parity, and edge expansion.
forecast · section 07 of 10
The 24-month horizon — six dated bets
Where do costs head, and which dated bets would falsify that?
Boundary
These are probability-bearing judgments over a short record, not calibrated forecasts; two of the six rest on structured priors with named missing baselines.
Source coordinate
Report heading The 24-month horizon — six dated bets, under What this means for decisions, in The real cost of AI: August 2026.
The section itself
The forecast notebook has now asserted its dated binary bets (evidence cut 2026-08-20, resolution by August 2028):
| Bet | Claim | P | Confidence |
|---|---|---|---|
| FC1 | Model efficiency (not hardware) drives >50% of further price decline | 0.70 | medium |
| FC2 | Hyperscaler custom silicon reaches 25%+ of inference workload by mid-2028 | 0.45 | low-medium |
| FC3 | Jevons paradox holds — a 50% unit-cost cut raises volume more than 50% | 0.65 | medium |
| FC4 | Subsidies contract 50%+, raising effective cost 1.5–3x for subsidized users | 0.50 | low-medium |
| FC5 | Open-weight models reach quality parity on most workloads within 24 months | 0.55 | medium-low |
| FC6 | Edge becomes cost-advantageous for a materially larger workload set | 0.60 | medium-low |
The structural read: the 10x/year compression era is ending (fixed-quality decline is decelerating toward 1.5–5x/year and bifurcating — commodity approaching free, frontier reasoning moving up in price), so planning should treat unit cost as a shrinking but non-zero line item while total spend likely still rises (FC3). The least evidenced bets — subsidy contraction and demand elasticity — are the ones that would move budgets most.
Who uses this section, and for what
Engineering leaders and CTOs
Six dated bets with probabilities, in place of a curve you cannot defend.
Platform and infrastructure teams
The hardware-cadence and custom-silicon bets that move the threshold under you.
Procurement and finance teams
The dated subsidy-contraction bet and its probability.
Founders and product teams
Which direction the inputs move over a funding cycle.
Executives, investors, and broad readers
Six dated bets, each with the condition that annuls it.
Reader questions that route through here
- Where are costs headed in 24 months?Answered as six dated bets
- Can I run inference on the edge?Answered in research custody; carried onto this site only as a forecast bet
- How do all these dimensions interact?Partially answered