---
id: "claim-efficiency-increases-demand"
type: "claim"
source_timestamps: ["§ The Great Value Loop", "¶9"]
tags: ["efficiency", "energy-demand"]
related: ["concept-ai-jevons-paradox", "contrarian-efficiency-increases-demand", "entity-deepseek"]
confidence: "high"
testable: true
sources: ["futures"]
sourceVaultSlug: "hbr-seg-futures"
originDay: 2
articleStem: "hbr-nm-101-energy-strategy-ai"
sourceUrl: "https://hbr.org/2026/06/your-company-needs-an-energy-strategy-for-ais-next-phase"
sourceTitle: "Your Company Needs an Energy Strategy for AI’s Next Phase"
---
# AI efficiency gains will increase total energy demand

## Claim
Because of the Jevons paradox, making AI models cheaper and more efficient to train and run (as demonstrated by [[entity-deepseek-d2]]'s 2025 release) will **not** reduce aggregate energy pressure. Instead, it will expand the number of economically viable use cases for AI, driving total energy demand upward.

**Confidence:** high · **Testable:** yes

This claim is the mechanism [[concept-ai-jevons-paradox]] in action, and its counterintuitive edge is captured in [[contrarian-efficiency-increases-demand]].

## Enrichment (external validation)
- **Americans for Prosperity:** forecasts where AI alone could drive up to a **165% increase in power demand by 2030**.
- **WEF:** AI data-center investment is outpacing grid build-out despite ongoing hardware/software efficiency improvements.
- **Brookings:** AI is also used to *increase* grid efficiency (potentially freeing ~175 GW of transmission capacity) — a double-edged pattern where efficiency in one part of the system enables more total usage elsewhere.

## Nuance
Expert restatement: *"Absent binding caps on compute or energy, AI efficiency improvements are likely to produce large rebound effects, increasing total energy demand."* Under strict caps or carbon pricing, efficiency could instead reduce total use.


## Related across articles
- [[concept-induced-demand]]
- [[contrarian-inefficiency-is-good]]