Altman Slams AI's Massive Power Use

#energy-consumption#sustainability#ai-criticismaisam-altman

💡Altman's energy critique highlights sustainability must for AI builders

⚡ 30-Second TL;DR

What Changed

AI training energy equals 20 years of human rice consumption for smarts

Why It Matters

Sparks debate on AI sustainability, urging practitioners to prioritize energy-efficient models amid growing infrastructure costs.

What To Do Next

Audit your model's carbon footprint with MLCO2 library to optimize inference energy.

Who should care:Founders & Product Leaders

Background and context from public sources — not the original article. 6 sources cited.

•Altman dismissed viral claims that ChatGPT uses multiple gallons of water per query as 'totally insane' and outdated, citing the shift away from evaporative cooling in data centers.[1][3]
•A single ChatGPT query consumes about 0.34 watt-hours of energy, equivalent to a lightbulb for a few minutes or an oven for one second.[4]
•Altman has invested in nuclear fusion (Helion) and small modular nuclear reactors (Oklo) to address AI's energy needs.[2]
•IEA reports data centers used 1.5% of global electricity in 2024, with AI-driven demand growing over 4x faster than overall electricity and projected to double by 2030.[2]

AI inference energy efficiency will surpass human levels by 2030

Altman states AI has already matched humans on per-query energy post-training, amid data center demand doubling per IEA forecasts.[2]

Nuclear power will supply over 20% of AI data center energy by 2030

Altman advocates rapid nuclear, wind, and solar expansion, backed by his investments in Oklo and Helion amid rising AI energy needs.[1][2]

2024-04

IEA reports data centers at 1.5% global electricity use with rapid AI-driven growth.

2025-06

Altman states average ChatGPT query uses 0.34 watt-hours on X.

2026-02

Altman speaks at Indian Express AI summit, defends AI energy use with human comparison and calls for nuclear shift.

Factual claims are grounded in the sources below. Forward-looking analysis is AI-generated interpretation.

Weekly AI Recap

Read this week's curated digest of top AI events →

Same topic

Explore #energy-consumption

Same product

AI-curated news aggregator. All content rights belong to original publishers.
Original source: Ifanr (爱范儿) ↗

This is a summary, not the original. Read the source, or get the weekly briefing.

Weekly AI briefing

One email a week. Unsubscribe anytime.