It looks like the “JVG algorithm” only wins on tiny numbers

Quantum “cheat code” gets roasted — only works on toy numbers

TLDR: Scott Aaronson shot down the hyped “JVG” quantum shortcut, saying it depends on precomputing an enormous list—fine for tiny demos, useless for real encryption. Most commenters applauded the smackdown, while a few defended small wins and naming conventions; many called the non‑arXiv preprint and clickbait coverage red flags.

The quantum crowd is feasting after computer scientist Scott Aaronson torched the so‑called JVG algorithm, a splashy claim that promised to beat the famed Shor’s algorithm and crack big encryption with just 5,000 qubits. The reveal? Their “trick” is to precompute a mind‑boggling list and shove it into a quantum machine—great for toy examples, totally hopeless for real‑world sizes. Aaronson’s tone? Spicy, calling the authors “intellectual hooligans,” and the comments cheered like it was open‑mic night at the roast.

The strongest vibe: this was hype on stilts. Folks flagged red‑alert signs—no arXiv paper, just a Preprints.org upload—and noted that only clickbait sites amplified it, while respectable science outlets steered clear. One user even tracked the top HN thread and a calmer PostQuantum explainer as the “polite versions” of the same facepalm.

But it wasn’t a total pile‑on. A few contrarians cracked that “hey, a speedup on tiny numbers is still something,” and another defended naming the algorithm after its authors, comparing it to AVL trees—not a crime. The more academic‑minded asked if it fits in BQP (translation: problems quantum computers can solve efficiently), but even they conceded that “precompute everything” usually means “wait forever.”

Meme watch: commenters riffed on Aaronson’s “any undergrad could spot this” line, joking that freshmen were being summoned with calculators. Verdict: the internet loves a flambé, and today’s special is a quantum shortcut that only sprints when the race is two steps long.

Key Points

  • The JVG algorithm claims to vastly improve on Shor’s factoring and to break RSA‑2048 with about 5,000 physical qubits.
  • Its core idea is to precompute all x^r mod N values classically and load them into a quantum superposition.
  • Aaronson argues this requires exponential time due to exponentially many r values, making it impractical beyond tiny inputs.
  • The paper appeared on Preprints.org rather than established repositories like arXiv, ECCC, or IACR, which is flagged as a red flag.
  • The claim was amplified by clickbait sites but largely ignored by reputable science news; external explanations are linked from Hacker News and Postquantum.com.

Hottest takes

“lambasting the paper… ‘intellectual hooligans’” — MathMonkeyMan
“a speedup on tiny numbers is still impressive :P” — kmeisthax
“The same way the AVL tree is named… Nothing peculiar” — guy4261
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.