Skip to content
SUPREME AI AGI

Prediction · On the road to AGI

Electricity and compute, more than ideas, will be the main limit on how fast AI capabilities grow over the next decade.

Medium termMedium confidenceOpen

Why we think so

The IEA projects data-centre electricity use roughly doubling to about 945 TWh by 2030, driven by AI; the AI Index reports AI data-centre capacity of 29.6 GW. Gains increasingly come from using more compute at the time of use, which adds to demand.

What would prove us wrong

Large efficiency gains letting capabilities keep rising while AI's electricity use flattens.

Horizon
Medium term. About 3 to 10 years out.
Confidence
Medium confidence
Status
Open. Not yet decided either way.
Made
21 October 2026

The evidence it rests on

Public results from teams around the world. Read the sources and judge for yourself.

  • · International Energy Agency

    IEA: data-centre electricity use set to roughly double by 2030

    Data centres used about 415 TWh in 2024. The IEA projects about 945 TWh by 2030, with AI the main driver, and notes large uncertainty.

    What it means for AGI: energy and compute are becoming physical limits on how fast AI can scale.

    Source: iea.org(opens external site)

  • · Stanford Institute for Human-Centered AI

    AI Index 2026: capabilities jump, but unevenly

    Agent success on Terminal-Bench rose from 20% to 77.3% in a year, and agents solved 93% of cybersecurity tasks, yet robots succeeded at only 12% of real household tasks. AI data-centre power capacity reached 29.6 GW, and four in five US students use AI for schoolwork.

    What it means for AGI: digital abilities are racing ahead while physical abilities and governance lag.

    Source: hai.stanford.edu(opens external site)

Think we are wrong?

Send us the evidence. If it changes our view, we update the prediction and say why.

Challenge this prediction