Is China beating the US to AGI?

The essay's thesis

A central pillar of Situational Awareness is a US–China race: as AGI approaches, algorithmic secrets and model weights become the decisive strategic asset, and the US must "lock down the labs" to convert its compute and talent lead into a durable advantage. The implicit prediction is that the gap stays wide enough to matter.

Where the race stands in mid-2026

The frontier is still American-led — but the margin is thin. Open-weight Chinese models compressed the gap to roughly 3–6 months at a fraction of the cost, with genuine architectural innovation (Epoch AI characterizes DeepSeek's Multi-head Latent Attention and fine-grained Mixture-of-Experts as real advances, not mere distillation). A months-wide, rapidly-closing gap is very different from the durable moat the essay's strategy assumes.

Why the narrow gap matters

This is the same evidence behind the scorecard's one Wrong verdict — open source did not fade. If capability diffuses cheaply across borders, compute concentration buys less geopolitical leverage than the essay's model assumes, and the case for a government "Project" weakens. So "is China beating the US?" and "does the US have a moat?" are really the same question — and on the moat, the answer so far is no.

What to watch

The verdict would shift back toward Aschenbrenner if export controls or compute costs push the open-weight gap back beyond ~18 months for two consecutive frontier generations. Until then: the US leads, but not durably.

Frequently asked questions

Not as of mid-2026 — the US frontier still leads. But open-weight Chinese models like DeepSeek V4 and Qwen 3.7 Max trail by only ~3–6 months at far lower cost, a much narrower gap than Aschenbrenner's framework assumed.

Not a durable one, so far. The lead is months, not years, and closing — which is why the scorecard grades Aschenbrenner's 'open source fades, proprietary moat holds' prediction as Wrong.

Roughly 3–6 months behind the proprietary frontier as of 2026, with genuine architectural innovation and dramatically lower pricing — not merely distillation of Western models.