Kindly Verify

The Word That Does Too Much Work

Dhirender

Three different stories used the same word: quantum. Two of them made news in the first week of October. The third has been running for more than two years. In one, two Indian institutions sent a secret message that nobody can secretly read, because of a strange rule in physics. In another, an Indian company raised money for its “quantum” chips, and the one hard number in the coverage was how briefly they hold their special state, far too short for most real jobs. In the third, an app called “Quantum AI” used fake videos of a famous billionaire to trick people into handing over money, in several countries, and had nothing to do with real quantum physics at all.

The word was the same in all three stories. What it was describing was not.

The One That’s Real

This part of the story comes from a press release that the Press Information Bureau (PIB), the government’s news office, put out on 3 October. It is genuinely worth reading in full.

First, what does “quantum” actually mean? Very tiny things, like a photon, a single speck of light, do not have to be one definite thing until someone checks. Imagine flipping a coin and catching it in your closed hand, before looking. Here the picture breaks, and the break is the point. A covered coin is already heads or tails. You just do not know which. A photon is not like that. Experiments have ruled out the simple idea that its answer was there all along. The moment you check it, it settles, and it cannot go back to being undecided.

This is what makes quantum key distribution special. Two people share a secret code by sending it as a stream of photons. If anyone else tries to read the code along the way, the act of reading changes the photons. The two people can tell, from the errors that show up, that someone looked. Nobody can secretly listen in without getting caught, not because the system was cleverly designed, but because of how these particles behave in nature.

According to the PIB release, BISAG-N and IIT Gandhinagar did exactly this on the night of 27-28 September, over a distance of 5.56 kilometres, through open air rather than a cable. QNu Labs, an Indian company, built the equipment, called Armos, that sent and caught the photons. The release reports that fewer than 5 out of every 100 bits of the shared code came through with an error, a number called the quantum bit error rate, and that a fresh, secure code was produced at a rate of 230 to 260 bits every second.

The release also describes a second layer, BISAG-N’s own software, called Vedic Kavach. It adds two things using ordinary computer code: post-quantum cryptography, a lock designed to stay hard to pick even for a future, more powerful quantum computer, and quantum random number generation, meaning the codes come from genuinely random quantum events rather than a formula a computer could predict. This layer keeps things working even if the photon link is briefly interrupted.

This is useful for protecting secrets that need to stay safe for years, in defence, banking and power systems, since data captured today could otherwise be stored and decrypted later as quantum computers improve. The release also frames the project as a step toward self-reliance, using indigenous technology from an Indian company and two Indian institutions.

One honest limit still applies. The physics cannot be fooled, but the equipment can. In 2010, researchers blinded the light detectors of two commercial quantum cryptography systems and copied the keys without being caught. This is still a real, working demonstration of real quantum physics, reported plainly by the people who did it.

The One That’s Real, But Not Yet

Quantum computing is a different claim, built on the same physics but much further from finished. Quantum key distribution only needs one photon to stay undecided for a moment, just long enough to cross 5.56 kilometres. A quantum computer needs many such particles to stay undecided together, for long enough to finish a whole calculation. How long they can stay that way is called coherence time, and it is the biggest thing standing between today’s machines and the glowing photo on a brochure.

Picture a spinning top. A quantum computer needs its particles spinning, like that top, undecided, for the whole time it is doing its sums. The moment a particle wobbles and falls over, settling into one definite answer too early, that part of the calculation is ruined.

QpiAI, an Indian company that builds quantum chips, raised ₹50 crore in debt in the first week of October. A TechTimes report on the deal gave a number for its machine: a target of about 30 millionths of a second for keeping its particles spinning, with a 2026 roadmap goal of 100. That sounds short because it is: computers reliable enough to beat ordinary computers at real problems need far longer.

Notice where that number came from: a report, not QpiAI’s own words. Still, it is a specific figure that anyone can check. The third story in this piece never offers one.

The One That Was Never Quantum At All

In 1994, in a book called The Quark and the Jaguar, the physicist Murray Gell-Mann, who won a Nobel Prize, made up a word for something he kept noticing: quantum flapdoodle. He meant people borrowing fancy quantum-sounding words to seem smart, without actually explaining anything real. He noticed this pattern decades before today’s online scams existed to use it.

“Quantum AI” is not one company. It is a network of fake investing apps using that name, or names close to it, across several countries, promising guaranteed profits from an AI it claims is powered by quantum computing. To convince people, it has used deepfakes, fake AI-made videos that make a real person appear to say or do something they never did, including ones making it look like a well-known billionaire recommends it.

Hong Kong’s financial regulator has publicly warned of suspected fraud under this name, and a UK consumer group, Which?, has separately exposed it as a global scam. In Australia, people reported losing more than A$8 million in 2023 to fake AI trading platforms of this kind, according to the national anti-scam centre run by the consumer watchdog, which thinks the true figure was closer to A$20 million.

India has seen it too. Earlier this year, PIB Fact Check warned that a “Quantum AI” investment scheme was being run in the Union Finance Minister’s name, and said that neither the minister nor the Government of India endorses any such scheme. In Maharashtra’s Palghar district, cyber police froze about ₹38 lakh of the roughly ₹48.5 lakh a retired bank employee allegedly lost to a fake “Quantum AI” platform. No quantum computer anywhere predicts stock markets. Nothing in the first two stories above, not QpiAI’s chip, not the BISAG-N link, is what any trading app could plausibly be running. Here, the word “quantum” was not describing anything real. It was simply borrowed, for its reputation.

What I believe, and cannot prove

Here I stop reporting. What follows is what I believe, not something I can prove.

I think the word “quantum” gets borrowed so easily because it is one of very few science words that sounds careful and magical at the same time. Words like “coherence time” or “post-quantum cryptography” sound exactly like what they are: precise, limited, something you could check. “Quantum” on its own sounds like it could mean almost anything. A real physicist and a scammer can use the exact same word.

I know the weak point in this belief. I am noticing a pattern from a handful of stories, not something measured over time, and I cannot prove “quantum” is specially vulnerable rather than just the current favourite in a much older habit, one that has also attached itself to “magnetic,” “atomic,” and “cyber.” Weigh this the way you would weigh anyone’s belief: by what it rests on, which here is three documented stories and my own guess about why they share a word.

What you can do on Tuesday

Next time you see the word “quantum” attached to something, try this test: is there a specific, checkable fact behind it, a coherence time, a key rate, a published study? Or is it just an adjective, like quantum energy, quantum frequency, quantum healing, with nothing solid underneath? If it is just the adjective, that missing detail is your answer. Real quantum science states its limits in numbers, the way the PIB release gave its error rate and its key rate. A word doing too much work is exactly what Gell-Mann named in 1994, and what watchdogs on three continents have had to call out in recent years.

October 20, 2026  |  dhirender.tamber@gmail.com

Sources

Press Information Bureau (Ministry of Electronics & IT), release on the BISAG-N–IIT Gandhinagar free-space QKD demonstration (3 October 2026); QNu Labs, press release on the same trial (5 October 2026); TechTimes, “QpiAI Wins Institutional Debt on 60% Margins: India Quantum Hardware Grows Up” (5 October 2026); YourStory, “Quantum computing and AI firm QpiAI secures Rs 50 Cr debt financing from InnoVen Capital” (5 October 2026); Lars Lydersen and others, “Hacking commercial quantum cryptography systems by tailored bright illumination,” Nature Photonics (2010); Royal Swedish Academy of Sciences, Nobel Prize in Physics 2022, popular science background; Murray Gell-Mann, The Quark and the Jaguar (1994), chapter “Quantum Mechanics and Flapdoodle”; Marin Ivezic, “Quantum of Flapdoodle: A Guide to Quantum Hype and Scams,” PostQuantum.com (published 29 January 2025, updated 7 September 2026); Hong Kong Securities and Futures Commission, public warning and alert-list entry on “Quantum AI” (8 May 2024); Which?, “We exposed a global AI scam” (26 November 2024); National Anti-Scam Centre (ACCC), Investment scam fusion cell final report (May 2024); PIB Fact Check, post on a fake “Quantum AI” scheme (24 May 2026); The420.in, “₹38.21 Lakh Frozen in Palghar Fake Quantum AI Investment Scam” (9 July 2026).