A design specification
Greener
Hours.
An open accountability standard for AI compute carbon disclosure — three implementation surfaces that turn invisible energy costs into legible procurement infrastructure.
§ 03 · The problem · the scale
Projected global data-center electricity demand by 2030 — more than double 2024's 415 TWh, with AI identified by the IEA as the lead driver.
IEA, Energy and AI (April 2025); Shehabi et al., 2024 U.S. Data Center Energy Usage Report, LBNL (Dec 2024).
§ 04 · The invisibility
The user sees the AI.
They don't see any of this.
Dark-matter framing after Dan Hill, Dark Matter and Trojan Horses (Strelka, 2015); AI energy estimates: IEA, Energy and AI (2025).
§ 05 · The reframe
Visibility doesn't change end-user behavior. But visibility paired with a standard creates supply-side pressure.
The same pattern as HTTPS, energy labels, SBOM, and cloud-carbon dashboards.
End users didn't change.
Providers did.
A backend protocol made visible at the UI layer through a tiny glyph. No one wanted their site flagged “Not Secure” in Chrome. Within a decade the entire web shifted.
The same pattern, across domains
The A–G grade reshaped manufacturer behavior more than consumer behavior — appliances were re-engineered to climb the visible scale.
Barely moved consumers; drove reformulation — trans fats, sodium, sugar — because makers competed on what was visible.
AWS, Azure, GCP shipped per-service carbon in 2021–22 on procurement demand. The pattern works; the AI-specific extension is what's missing.
Google Transparency Report; Let's Encrypt annual reports (2017–24). Curve directional. Inflection: Chrome 68 (July 2018) flagged all non-HTTPS pages.
§ 06 · The opportunity · three forces
Why now, not five years ago.
Regulatory pressure
All demand vendor-level emissions data. None specify the technical standard.
Procurement demand
Cloud got per-service carbon dashboards. AI didn't.
Reusable infrastructure
Three layers already exist. The fourth is the work.
§ 07 · The product · the standard
One standard. Three surfaces.
Open-source spec · reference SDK in Python, TypeScript & Go · piggybacks on existing API plumbing. Adoption is three headers.
§ 08 · The three surfaces
One indicator, one scheduler, one dashboard.
Compute Window Indicator
A small glyph that makes invisible energy costs visible — without pretending it changes user behavior.
The P-codes on each surface map to the ten design principles from Prof. Godelnik's Design for a Warming World — the project was developed roughly against that framework.
§ 15 · Named, not hidden
What this project cannot pretend.
The structural tension
The biggest lever for AI's footprint is supply-side — siting, PPAs, grid decarbonization. Greener Hours operates on a derivative lever: user-facing legibility pressuring the supply side. The bet is contestable; the HTTPS analog suggests the pattern works, but the timeline is uncertain.
Greenwashing capture risk
Open standards get captured. Three guardrails: an open-source reference implementation, third-party verification of disclosure claims, and an anti-rebound layer that surfaces total consumption alongside efficiency. None is sufficient alone.
Visibility ≠ behavior change
The sharpest midterm critique, addressed structurally: Tier 1 is not a behavior tool. Behavior change happens at Tier 2 (a one-click default) and Tier 3 (institutional decisions). Tier 1's job is legibility that creates supply-side pressure.
The sufficiency limit
Greener Hours is a timing-and-disclosure intervention, not a sufficiency one. It does not advocate using less AI. Tier 3 makes absolute usage visible — but visibility is not advocacy.
§ 16 · If the wedge works
When compute disclosure becomes a normal procurement field.
major AI providers implement the spec, on competitive and procurement pressure.
of enterprise AI API calls return standardized carbon data by default.
of deferrable batch tasks shift to cleaner grid windows.
Fortune 100 buyers name the spec in AI RFPs.
EU AI Act or California SB 253 references the spec — the mandate moment.
“What's your compute-carbon disclosure?” becomes a normal sales question.
Directional, not predictive — what a working wedge could move, not measured outcomes.
§ 17 · Close
Make the invisible legible, and the providers move.
The three surfaces above are live — on one simulation.
Tab through them: pause the scheduler's clock and the chat's carbon indicator freezes with it; submit a flexible job and the dashboard's flex strip picks it up. The chat runs on a real model API through a server-side proxy, grounded in the sim state it sits on. Greener Hours itself remains a speculative open standard — the full argument lives in the presentation deck, linked above.
Open the prototype full-screen