Is AGI inevitable?

The case that it's close to inevitable

The engines keep running. Effective compute scales ~0.5 orders of magnitude per year, capex has exceeded even aggressive projections, and capability keeps climbing (~83% GDPval, ~80% SWE-Bench Pro). Expert medians have compressed from ~2060 to ~2033 in six years — the whole distribution keeps sliding earlier. When the inputs compound and the forecasts converge, "it won't happen" gets hard to defend.

The case that it isn't guaranteed

Two real off-ramps remain:

• The unproven step. AGI requires reliable, autonomous, end-to-end work — and no system has demonstrated it. Capability has stalled at exactly that line before; there's no law that says it must cross.
• The trends must hold. Compute scaling depends on capex, and capex depends on revenue eventually justifying it. A funding pullback — the bear case behind the capex-bubble question — would slow everything.

So "inevitable" is the wrong word. "On track, with identified ways it could fail" is the honest one — which is exactly how this scorecard grades the AGI-2027 prediction: Open, not certain.

How to tell trending from certain

Watch the two things that would settle it: whether the autonomy step gets crossed, and whether the input trends hold. Both feed the AGI-2027 Thesis Tracker — one auditable score that captures how the whole bet is holding up, moving only on evidence. The headline claim resolves by January 1, 2028.

Frequently asked questions

Strongly trending, but not guaranteed. The inputs — compute (~0.5 OOM/yr), capex, capability — keep climbing and expert timelines keep moving earlier, so almost no serious forecaster says 'never.' But AGI still depends on crossing one unproven step (reliable autonomous work) and on those trends continuing, so 'inevitable' overstates it.

Most evidence points toward yes eventually, but 'definitely' is too strong. Two off-ramps remain: the autonomy step (reliable unsupervised end-to-end work) is undemonstrated, and the compute/capex trends driving progress could slow if funding pulls back.

It's possible but looks unlikely on current trends. The main ways it stalls are capability plateauing at the autonomy line, or a capex/funding pullback slowing compute scaling. Neither is happening as of mid-2026, but neither is ruled out.

That's the optimists' framing, and the trends lend it support — but the honest read is 'on track with identified failure modes,' not certainty. The AGI-2027 verdict is graded Open, not guaranteed; it resolves by January 2028.