Quantum Computing: Is It Bullshit? (The Honest Answer)

The biggest number a quantum computer has ever factored using Shor's algorithm, the method that would break today's encryption, is 21. Three times seven. My kid could do that at three. Yet governments and companies are spending billions to protect themselves from these machines, and you'll hear that the internet breaks in 2029. So which is it? Is quantum computing a real threat, or is it bullshit?

I've spent 30 years in cybersecurity and I've watched five cryptographic panics come and go. This is the sixth. Most of them turned out to be half true. So I went past the press releases and into the research itself, the papers cryptographers and physicists are actually arguing about. Here's how I think this one plays out.

In this guide, we'll look at the evidence on both sides, a realistic timeline, why you still need post quantum cryptography even if the skeptics are right, and the specific moves to make now. Everything has been checked again against the sources on 1 October 2026.

Cryptocurrency vs Cryptography: The Confusion Driving the Panic

A few months ago, YouTubers told tens of millions of people that quantum computers were about to kill "crypto". They cited Scott Aaronson, the quantum computing researcher best known for puncturing hype, who wrote in April 2026 that experts he trusts think a code-breaking machine "ought to be possible by around 2029". They cited Google's 2029 plan. And they cited a new paper estimating that about 26,000 physical qubits could break the kind of encryption Bitcoin uses.

Before we look at any of that, one clarification, because it's doing a lot of the work in this panic. "Crypto" means two completely different things.

Cryptocurrency is Bitcoin and its relatives. Cryptography is the maths that protects every password, every bank transfer and every encrypted message you send. It's what makes buying something online safe. And not all of it is at risk. The part quantum computers threaten is public-key cryptography: the maths that lets two strangers agree a secret key, and that proves who signed something. Its security rests on hard maths problems, factoring huge numbers (RSA) and a related problem called the discrete logarithm (elliptic curves, which Bitcoin uses). Shor's algorithm, run on a big enough quantum computer, solves both. Hash functions and the symmetric encryption that protects data once a key is agreed are far less affected. Thumbnails say "crypto" and mean the coins; the videos often switch to the maths without saying so. Both are in scope, because Bitcoin is built on cryptography too, but they're very different problems. We'll come back to Bitcoin at the end.

What Google's 2029 Date Actually Means

Almost nobody repeating the 2029 date has read what Google wrote. In a March 2026 blog post, Google set itself a deadline: finish moving its own products and infrastructure to post quantum cryptography by 2029. That's a migration target. It isn't a prediction that a working code-breaking quantum computer will exist by then, and nobody at Google said it was.

Those are two completely different statements. One says "we want to be ready early". The other says "the internet breaks in 2029". If you read the headlines, you'd think it was the second.

The Skeptic Case: 21 Is Still the Record

So what does the evidence actually look like? Start with the record. The first quantum factorisation, in 2001, factored 15. Not a 15-digit number: the number 15, which is three times five. The next record, in 2012, was 21. A 2019 attempt at 35 failed. As of October 2026, there's no new genuine record.

It's worse than it sounds. Even those demos used what's called compiled Shor's algorithm: the researchers built knowledge of the answer into the setup. That's not factoring, it's confirming. Bigger "records" you may have seen in press releases relied on similar shortcuts.

The cryptographer Peter Gutmann of the University of Auckland made the point brilliantly. In a 2025 paper with Stephan Neuhaus, Replication of Quantum Factorisation Records with an 8-bit Home Computer, an Abacus, and a Dog, they matched every record with a 1981 home computer, an abacus, and a dog trained to bark three times. Bruce Schneier, one of the most respected cryptographers alive, wrote it up approvingly. Gutmann's talk on the subject is bluntly titled "Why Quantum Cryptanalysis is Bollocks", and his summary line is: "Quantum cryptanalysis is the string theory of security." Lots of theory, almost no data.

That's the skeptic case. Hold it in your head, because the other side has evidence too.

The Other Side: Error Correction Is Working

In my opinion, the skeptics aren't entirely right either. The big obstacle to useful quantum computers is noise: qubits make errors. The fix is error correction, which combines many noisy physical qubits into one reliable "logical" qubit. For years nobody could show that adding more qubits actually reduced errors. Now they can.

Google showed a bigger error-correcting code beating a smaller one in a 2023 Nature paper, though only by about 4%. Its Willow chip went much further in December 2024: each step up in code size roughly halved the error rate. That's the first strong evidence that scaling works.

Logical qubit counts are climbing too. In the video I quoted 12, from Microsoft and Quantinuum in 2024. Quantinuum's Helios machine has since run 48 error-corrected logical qubits.

Now compare that with what's needed. The best current estimate for breaking RSA-2048, the encryption behind much of the internet, is about 1,400 logical qubits and fewer than a million physical ones (Craig Gidney, Google, 2025). For the elliptic-curve maths Bitcoin uses, Google researchers estimated 1,200 to 1,450 logical qubits in March 2026. So 48 isn't most of the way there. It's a few percent of the way there. But it's moving in the right direction, and the estimates of what's needed keep falling: the Oratomic and Caltech paper behind the "26,000 qubits" headline counts physical neutral-atom qubits, far fewer than the roughly one million physical qubits assumed in earlier estimates, though it's still a design on paper. (Physical and logical counts can't be compared directly: one logical qubit is built from many physical ones.)

Credentialed scientists still argue it can never scale. The mathematician Gil Kalai of the Hebrew University argues that correlated noise will stop error correction from ever reaching the level needed. His position hasn't been refuted. It's contested.

So here's where to land. One side has receipts that scaling works at small sizes. The other argues it will never scale to break RSA. Both have evidence, and neither has won. That makes a code-breaking quantum computer an engineering problem with an unknown completion date: not imminent, but not quite bollocks either.

Press Release vs Paper: Microsoft's Majorana 1

Here's a perfect example of the gap you're being sold. In February 2025, Microsoft announced its Majorana 1 chip and said it had created "an entirely new state of matter". Millions of views followed. Then people read the paper.

The peer-reviewed Nature paper was far more careful. It said its measurements couldn't, by themselves, show the exotic states the press release was celebrating. Press release said breakthrough. Paper said maybe. The argument has continued with Microsoft's Majorana 2 announcement in 2026.

That's the pattern. Quantum sells. Calm doesn't. One camp says a code-breaking machine is impossible, the other says it's imminent, and in my view both are wrong. The evidence is thinner than the panic.

Why You Still Need Post Quantum Cryptography Anyway

So why is everyone, including me, still telling you to swap today's public-key encryption for post quantum cryptography? Because it's insurance against a plausible threat, even if progress is as slow as the skeptics say. If Gutmann's "never" turns out to be right, the insurance wasn't needed. But you can't wait to find out.

There's a simple test called Mosca's inequality, after the Canadian researcher Michele Mosca. It says: if X plus Y is greater than Z, you have a problem today.

Twenty plus fifteen is 35. If Z is 6, or even 14, you're already exposed. Not because we know the date, but because the maths doesn't care whether we do. And attackers don't need the quantum computer today to cause the damage. The US government's June 2026 executive order on post-quantum cryptography names the risk of "adversaries collecting United States information now, and decrypting it later". That's harvest now, decrypt later: steal the encrypted data today, open it when the machine arrives.

This rhymes with Y2K, but it isn't identical. With Y2K we knew the date. Here we're planning a 10-year migration against a threat with an unknown completion date. The serious question isn't "when does quantum break crypto?" It's "how do we migrate in time without knowing?" The answer is: start anyway.

Four Wildcards Nobody Is Pricing In

Most people predicting quantum timelines ignore four things that could move them.

A Realistic Quantum Timeline

So here's what I'd remember. Yes, quantum computers will probably break today's public-key cryptography. Most likely between 2032 and 2040. Possibly later, and possibly never. The 2029 panic is wrong, and so are the deniers. The window has likely tightened because of AI and better designs. Tighter than the skeptics admit, slower than the panic merchants claim.

That's not just my read. In the Global Risk Institute's latest expert survey, published in March 2026, experts put the chance of a code-breaking quantum computer within about 10 years at 28 to 49%, and within about 15 years at 51 to 70%. And anyone who tells you the exact year, Altman, Aaronson, Gutmann, other YouTubers, me, is overconfident.

3 Post Quantum Cryptography Moves to Make Now (and 4 Things to Ignore)

Your job is to migrate calmly to hybrid systems, whichever extreme turns out to be true. Three things to actually do:

The deadlines are already set. You don't have to guess at Z to have a schedule:

And four things you can safely ignore:

If you're thinking about a career in this, post quantum migration is going to be a decade of work for security teams. It sits squarely in the skills I covered in I Scored 3,500 Jobs. These 6 Cyber Skills Win., especially security architecture.

Bitcoin and Quantum Computers: What's Actually at Risk

A specific word for Bitcoin holders, because I know you care about this. Bitcoin doesn't use RSA. It uses elliptic-curve signatures. Shor's algorithm breaks those too: different lock, same lock pick. But the risk isn't evenly spread.

Mining is fine for now. It relies on the SHA-256 hash function, which only gets a small quantum speed-up, and Bitcoin's difficulty adjustment absorbs it. The exposure is in coins whose public key is already visible on the blockchain. Google researchers estimate that's about 6.9 million BTC, roughly a third of all bitcoin, including about 1.7 million in early addresses from Satoshi's era. Other estimates range from 4 to 10 million. Once a quantum computer can run Shor's at scale, those wallets can be drained.

So if you hold Bitcoin, don't reuse addresses, and move coins from old addresses to a fresh, modern wallet well before a machine like that exists. Why does moving help? Modern Bitcoin addresses show only a hash (a scrambled fingerprint) of the public key until you spend from them, so an unused modern address doesn't expose the key a quantum computer would need. That reduces your exposure, though it isn't fully quantum-proof. Don't sell over a YouTube video, either. The protocol can probably be upgraded: there's a proposal, BIP-360, to add quantum-resistant outputs, but it's still a draft, and the upgrade politics could get ugly.

Y2K wasn't fake. The panic was wrong, and the migration was right. This rhymes with it: ignore the noise, migrate calmly, not because we know the answer, but because we don't. And quantum is only one piece of a bigger shift. AI is now finding vulnerabilities on its own, which I cover in The Future of Cybersecurity Just Changed.

Send me the book →

By submitting you'll get the book + occasional related emails. Unsubscribe anytime. Privacy.

Frequently Asked Questions

About the Author

Nathan House, Founder & CEO of StationX

Nathan House has 30 years of hands-on cybersecurity experience and is Cambridge-educated, holding CISSP, CISA, CISM, OSCP, CEH, and SABSA. He founded StationX in 1999 — one of the UK’s first cybersecurity companies — and has secured £71 billion in UK mobile banking transactions and the London 2012 Olympics, advising clients including Microsoft, Cisco, BP, Vodafone, and VISA. He authored the world’s most popular cybersecurity course — a #1 Udemy bestseller taken by over 500,000 students — and was named Cyber Security Educator of the Year 2020, AI Security Educator of the Year, and a UK Top 25 Security Influencer 2025. A DEF CON speaker and featured expert on CNN, Fox News, NBC, and the BBC, Nathan leads StationX’s training of more than half a million students worldwide.