A new preprint says near-term quantum experiments are chasing the wrong boundary

Amit Hagar's preprint, titled "The NISQ Trap", argues the past eight years of "quantum supremacy" experiments have been chasing the wrong boundary. NISQ, shorthand for Noisy Intermediate-Scale Quantum, names the 2018-present interim research phase between proof-of-principle chips and full fault tolerance. Quantum supremacy, the term the field has since softened to "quantum advantage," is the class of experiments claiming to do a specific task classical machines cannot efficiently reproduce. Hagar's specific claim: the regime quantum hardware can run with usable fidelity is the regime classical algorithms now compress.

Hagar backs that claim with six theoretical results from 2024 through April 2026, each arguing that some part of the circuit-space near-term hardware can reach overlaps with parts classical methods compress efficiently. The list is the load-bearing part of the paper. Read individually, the six results are claims about specific circuit families. Read as a block, they are Hagar's case that the pattern is not an accident.

Scott Aaronson, the UT Austin theorist who helped design and interpret several of the original sampling-based supremacy experiments, has published a public counter-essay titled "NISQ and quantum supremacy did not fail". His response is narrower than the framing suggests. He does not dispute that classical methods have caught up to specific demos. He disputes the move from "later classically reproduced" to "therefore proved nothing." A demo that ran on hardware classical methods could not match at the time, in the regime the experiment was specified for, did what it was meant to do. The fact that the regime was later compressed is a different claim, one Aaronson treats as separately falsifiable.

The boundary both sides are arguing about shows up in a peer-reviewed Nature paper from 2025, "Observation of constructive interference at the edge of quantum ergodicity". The paper studies the transition between quantum systems whose dynamics classical methods can simulate and systems whose dynamics they cannot, an "edge" Aaronson and Hagar would both agree is the right place to look. Where they disagree is whether the supremacy demos of the NISQ era actually probed that edge, or sat on the classical side of it.

There is no third voice in the public record. The preprint is not peer-reviewed, and the SciRate discussion thread on Hagar's paper shows community skepticism that goes beyond technical objections, including concerns about substantial AI generation of the manuscript. That is a separate story, but it shapes how much weight the six-results list can carry on its own.

If Hagar is right, the next decisive artifact is a quantum advantage demo that runs in a circuit regime the current compression theorems do not cover, with a classical hardness argument that pins down why that regime is hard and not just currently hard. If Aaronson is right, the next decisive artifact is a clean statement of what the existing demos did establish, a separation in the regime they ran in even if that regime was later compressed, and a clear accounting of what fault tolerance still has to add that NISQ never claimed to. A non-expert watching the next twelve to eighteen months should look for theorems that extend or close the specific circuit families Hagar's six results cover, and for experiments that target the constructive-interference edge in the Nature paper rather than the regimes the original supremacy demos used.

The two positions are not symmetric. Hagar is making an empirical claim that with one exception every NISQ-era flagship has been reproduced, reduced, or theorem-closed within roughly 18 months. Aaronson is making a methodological claim: that "reproduced" is the wrong verb. The next paper that runs the experiment Aaronson describes, or proves the theorem Hagar describes, is the one that will actually move the line.