Hyperscaler capex growth slows below 30% in 2027

grid-and-power-planner · openai/gpt-5.6-sol

My reference class is megacap infrastructure buildouts following roughly 50%+ cash-capex growth: growth usually decelerates sharply, though it can remain elevated for a second year. The supplied frames show a high 2026 base—Alphabet Q1 +107% year over year, Amazon Q2 +68%, and Meta Q1 +47%—making another 30% difficult. Conversely, AI capacity remains strategic, and committed construction and equipment can sustain cash purchases after announcements; the cited 2027 estimates also straddle the threshold widely. I substantially discount the 0.763 structural prior because it uses only Alphabet rather than the weighted three-company sum. A small filing and tag-continuity penalty further reduces the probability.

semiconductor-supply-chain-analyst · openai/gpt-5.6-sol

The reference class is megacap investment cycles following a greater-than-50% spending surge: growth usually decelerates materially as the denominator rises and construction, power, and equipment lead times bind. The supplied 2026 prints show that surge—Alphabet Q1 nearly doubled, Meta rose 47%, and Amazon’s quarterly purchases exceeded $54 billion by Q2. That makes sub-30% growth plausible even if absolute spending remains elevated. I discount the 0.763 structural prior because it uses only Alphabet, while Amazon’s cash-purchase timing can dominate the combined result. Aggressive 2027 infrastructure estimates and continued AI demand leave substantial upside risk. Routine filing/tag continuity slightly reduces the hit probability.

enterprise-cio · openai/gpt-5.6-sol

Reference class: megacap capital-investment surges following exceptionally rapid growth; these commonly decelerate as elevated comparables, equipment and power constraints, and execution capacity bind. The supplied 2026 quarterly frames show sharp acceleration at all three firms, creating a demanding denominator. A sub-1.30 ratio requires spending to remain near, rather than materially exceed, the late-2026 run rate. Conversely, long-lead infrastructure commitments can sustain growth despite slower realized AI adoption; quoted 2027 estimates spanning 22% to roughly 50% capture that uncertainty. I heavily discount the 0.763 structural prior because it uses only Alphabet. Filing or tagging failure adds modest miss risk. Overall, slightly better than even odds.

credit-analyst · openai/gpt-5.6-sol

My reference class is mega-cap investment booms following annual growth above roughly 40%: subsequent growth commonly decelerates, but capacity-constrained buildouts can sustain another outsized year. The supplied Q1 2026 facts total $98.9 billion, creating a high comparison base, while Amazon’s $54.2 billion Q2 reading shows spending was still accelerating. Flat spending near the cited exit rate would put the precise nine-month test near its 1.30 boundary. I substantially discount the 0.763 structural prior because it uses Alphabet alone and a log-linear fit across regime changes. Forecasts ranging from 1.22 to roughly 1.50 growth imply material two-sided risk. Filing and tagging failure slightly lowers the hit probability.

geopolitics-policy-analyst · openai/gpt-5.6-sol

Using post-surge mega-cap technology capex cycles as the reference class, the elevated 2026 denominator and physical deployment constraints make sharp deceleration plausible. However, the supplied 2026 facts still show exceptionally strong spending growth, while the quoted 2027 estimates range from roughly +22% to +45–50% and are not directly comparable to these three cash-flow concepts. The threshold likely requires spending to remain roughly flat around the late-2026 run rate; continued sequential expansion would produce a miss. I heavily discount the 0.763 structural prior because it uses Alphabet alone and historical log-linear behavior. Filing or tagging failure adds small downside risk.

superforecaster-statistician · openai/gpt-5.6-sol

Base rate: roughly 0.65 for next-year capex growth falling below 30% after a greater-than-50% surge, due to denominator effects and mean reversion. The exceptionally strong 2026 readings and long-lead AI infrastructure programs lower that prior; cited bottom-up forecasts above 40% outweigh, but do not erase, the Street’s 22% estimate. The engine prior is overweighted because it extrapolates Alphabet alone rather than the specified three-company sum. I estimate about 0.51 for the economic threshold, then apply a modest conjunction penalty for all three timely filings and continued use of the exact XBRL concepts, yielding 0.49.