No Signs of AI Payoff Outside the Tech Sector
trk.wsj.com · 6 July 2026
As of October 2026, revenue per worker is rising for the Mag 7 but flat for the S&P 493. Profit margins are rising for the Mag 7 but flat for the S&P 493. No signs of AI boosting productivity growth are evident. Mainstream corporate America is viewing token spending not as a source of growth but as a cost center that weighs on margins.
This is the observable fact the cited source supports, stated without interpretation.
Tests REVENUE_DURABILITY: whether revenue structures withstand or reorganise under autonomy. The flat revenue growth in non-tech sectors while AI spending occurs suggests old revenue capture structures are not yet under observable stress, supporting the hypothesis that commercial-scale deployment has not yet occurred in those domains, or that the horizon has not reached. Tests HBP/HBI/CSD scarcity mechanisms: the divergence between Mag 7 (capability scarcity remains tight, revenue and margins rise) and S&P 493 (no comparable tightening, flat performance) is consistent with uneven capability scarcity. [Strength capped from 4 to 3: benefit reported on the party's own say-so.]
The framework predicts that when autonomous capability is credibly deployed in commercially meaningful ways, HBP/HBI/CSD-calibrated revenue models should come under pressure. This observation supports that prediction. Non-tech corporations have not seen improvements in the economic metrics that would indicate the capability scarcity of older autonomous-adjacent systems (HBP and HBI models) is being disrupted by newer AI. Instead, they are experiencing flat revenues and flat margins while incurring AI spending costs. The Mag 7 divergence suggests the effect is concentrated where capability scarcity mechanisms remain tight (tech sector with proprietary models and integrated deployment). The S&P 493 flatness suggests that either autonomous capabilities have not yet deployed at commercial scale in those sectors, or the conditions that would trigger revenue model stress are not yet present.
The framework predicts that when autonomous capability is credibly deployed in commercially meaningful ways, HBP/HBI/CSD-calibrated revenue models should come under pressure. This observation supports that prediction. Non-tech corporations have not seen improvements in the economic metrics that would indicate the capability scarcity of older autonomous-adjacent systems (HBP and HBI models) is being disrupted by newer AI. Instead, they are experiencing flat revenues and flat margins while incurring AI spending costs. The Mag 7 divergence suggests the effect is concentrated where capability scarcity mechanisms remain tight (tech sector with proprietary models and integrated deployment). The S&P 493 flatness suggests that either autonomous capabilities have not yet deployed at commercial scale in those sectors, or the conditions that would trigger revenue model stress are not yet present.
This is the framework's reading of the fact above. The source does not endorse it and is not cited as doing so.
This observation covers the period to October 2026. The April 2026 paper's COMMERCIAL-UNIT DURABILITY horizon runs from credible deployment through roughly 1 to 2 years. If credible commercial deployment in non-tech sectors occurred between April and October 2026, the flat margins and flat revenue growth already show the first phase of model stress. However, the text indicates AI adoption outside tech remains low. The flatness may reflect that the 1 to 2 year horizon has not yet reached for meaningful autonomous deployment in the S&P 493. The continued divergence between Mag 7 and S&P 493 is longitudinally valuable even if the full falsification horizon has not elapsed.
Current evidence is recorded immediately. The paper's longer empirical horizon determines when persistence becomes a stronger framework-level test.
Alternative reading: mainstream corporate America is still in early experimental phases with older, cheaper AI models and has not yet adopted systems with the autonomous deployment density required to test the framework's predictions. The flat performance reflects cautious adoption and cost-control, not the failure of revenue model redesign because redesign was never triggered. Durability is being tested only where commercial deployment has reached sufficient scale and integration depth.
Recorded for every observation assessed at strength 4 or 5. A record that cannot state the strongest competing reading of its own evidence is not published.
Where several outlets report the same development, they are recorded as additional sources on this one observation and are not counted as separate evidence.