Skip to content
SUPREME AI AGI

Prediction · On the road to AGI

AI will match people at learning genuinely new tasks from a few examples, at a similar cost, which we see as one of the last clear gaps before AGI.

Long termLow confidenceOn track

Why we think so

Benchmarks such as ARC-AGI show that systems strong on exams still struggle to learn new rules from few examples. Narrow breakthroughs like AlphaFold show huge progress is possible once a problem is well framed; the open question is doing it without a person framing each problem.

What would prove us wrong

Systems matching people on generalisation tests only by spending far more compute than a person would need, or the gap staying open while other abilities grow.

Review notes

21 October 2026: Much faster than our long-term horizon: ARC-AGI-3 went from 0.51% at launch (March 2026) to 99.9% (September 2026) with a memory-preserving harness. Not marked happened: the standard-harness score was 62.7%, and the run cost about $18,800, far above what a person needs.

Horizon
Long term. More than 10 years out, or no clear timeline yet.
Confidence
Low confidence
Status
On track. New evidence points the way we predicted.
Made
21 October 2026
Last reviewed
21 October 2026

The evidence it rests on

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

  • · ARC Prize Foundation

    ARC Prize: generalisation is still hard for AI

    A public competition built on the ARC-AGI puzzles, which most people solve easily but AI systems find hard, to reward progress on learning new skills from few examples.

    What it meant for AGI in 2024: strong exam scores did not mean general intelligence; learning new tasks efficiently was the clearest open gap. See ARC-AGI-3 in 2026 for how that changed.

    Source: arcprize.org(opens external site)

  • · Google DeepMind (published in Nature)

    AlphaFold predicts protein structures at near-experimental accuracy

    A deep-learning system predicted the 3D shape of proteins with accuracy close to lab experiments, solving a problem biology had worked on for about 50 years.

    What it means for AGI: AI can make real scientific progress in a narrow field. The open question is whether the same can happen across many fields at once.

    Source: nature.com(opens external site)

  • · ARC Prize Foundation

    ARC-AGI-3 launches: people score 100%, frontier AI under 1%

    A new interactive benchmark of game-like environments with no instructions, rules or stated goals. At launch people scored 100% and the best frontier AI 0.51%.

    What it means for AGI: it tests exploring and learning in a new situation, the skill most clearly missing from AI at the start of 2026.

    Source: arcprize.org(opens external site)

  • · ARC Prize Foundation (testing OpenAI's model)

    GPT-6 Astra scores 99.9% on ARC-AGI-3, with a caveat

    ARC Prize verified 62.7% with its standard harness and 99.9% with a harness that keeps the model's reasoning state between steps; in one setting it used fewer actions than the median person on 96% of levels. ARC Prize states it is not claiming this is AGI.

    What it means for AGI: the gap on learning new tasks closed within six months of the test's launch. The jump from 62.7% to 99.9% came from how the system keeps memory, not from a new model, which supports the view that AGI is a system, not one model. The run cost about $18,800, far more than a person would.

    Source: arcprize.org(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