The Article That Started It
I read a breathless headline in November: "Quantum Computing Will Break Encryption by 2028." My first reaction was mild panic. My second reaction, about ten seconds later, was skepticism. Headlines about quantum computing have been promising imminent revolution for fifteen years. I wanted to understand what had actually changed versus what was still marketing.
So I spent a week going down the rabbit hole. Not as a physicist — I'm a software engineer with a rusty college physics background — but as someone who wanted to separate the real from the hyped. I read the NIST post-quantum cryptography standards. I watched IBM's quantum computing lectures on YouTube. I ran a Bell-state experiment on IBM's quantum cloud using fifteen lines of Python. I talked to two friends who work in the field.
Here's what I learned, in plain language, with the caveats I wish every article about quantum computing included.
The One-Sentence Version
Classical computers store information as bits — zeros and ones. Quantum computers use qubits, which can exist in a blend of zero and one simultaneously, and can be linked to other qubits in ways that have no classical equivalent. This lets quantum computers explore certain problem spaces in ways that classical computers can't match.
That's the honest one-sentence version. Everything after this sentence is nuance, and the nuance matters more than the headline.
Why Qubits Are Weird (and Why That Matters)
The weirdness of qubits comes down to two properties: superposition and entanglement.
Superposition means a qubit can be in a combination of zero and one at the same time. Not "it's zero sometimes and one other times" — it's literally in both states simultaneously until you measure it. When you measure a qubit, it collapses into either zero or one, and the probability of each depends on how the qubit was prepared.
Entanglement means two qubits can be linked so that measuring one instantly tells you something about the other, regardless of distance. If two qubits are entangled and you measure one as zero, the other will be one (or zero, depending on the entanglement type). This isn't communication — you can't use it to send messages faster than light — but it does mean that entangled qubits can represent correlations that classical bits simply can't.
Here's why this matters for computing: with classical bits, a computer exploring a problem with a trillion possible answers has to check them one at a time (or use enormous parallel hardware). A quantum computer with enough qubits can, for certain problem structures, represent all those possibilities simultaneously and use interference — the wave-like behavior of qubits — to amplify the right answers and cancel the wrong ones.
The key phrase is "certain problem structures." Quantum computing is not a faster classical computer. It's a fundamentally different tool that excels at specific types of problems and is useless at most others. Your email, your spreadsheets, your web browsing — none of these benefit from quantum computing. If someone tells you their quantum-powered app will make your daily life faster, they're selling something.
What Actually Changed Recently
Quantum computing has been "five years away" for about twenty years. What changed in the last two years that made the field genuinely exciting?
The answer is error correction. Qubits are absurdly fragile. Any interaction with the environment — a vibration, a temperature fluctuation, a stray electromagnetic field — can destroy the quantum state. This is called decoherence, and it's the central problem of quantum engineering.
For years, the solution was brute force: use thousands of noisy physical qubits to create one reliable "logical" qubit through error correction. The ratio of physical to logical qubits was enormous — sometimes thousands to one. Building a useful quantum computer at that ratio would require millions of physical qubits, which was engineering-wise impossible.
In late 2024, Google demonstrated something researchers had chased for decades: when you increase the size of the error-correcting code, the logical qubit actually gets more reliable. More physical qubits per logical qubit means fewer errors, not more. This sounds obvious, but it wasn't guaranteed. It's the difference between "quantum error correction is theoretically possible" and "quantum error correction actually works in practice."
Since then, the improvements have been steady. Logical qubit counts are up. Error rates are down. The honest metric for quantum computing progress is no longer "how many physical qubits" but "how many reliable logical qubits." That shift is the real news.
Where Quantum Computing Actually Helps
There are three domains where quantum computing offers genuine advantages, and it's worth being specific about each.
Simulating molecules. Molecules are quantum objects. Their behavior is governed by quantum mechanics. Classical computers approximate molecular behavior with brute-force calculations that grow exponentially with molecule size. A quantum computer simulates molecules natively. This matters for drug discovery — predicting how a candidate molecule interacts with a protein — and for materials science — designing better batteries, catalysts, and superconductors.
Drug companies are already paying for quantum cloud access to run molecular simulations. The results aren't replacing wet-lab experiments yet, but they're narrowing the search space enough to save months of development time. This is the most commercially advanced quantum application today.
Optimization. Routing delivery trucks, scheduling airline crews, balancing supply chains — these are combinatorial problems where the solution space explodes with size. Quantum approaches can sometimes find better solutions than classical heuristics, especially when combined with classical optimizers in hybrid workflows.
The honest caveat: the evidence here is still emerging. Quantum advantage for optimization is not proven in general. It works for specific problem structures, and identifying those structures is itself a hard problem. Anyone selling "quantum optimization" as a plug-and-play solution is overselling.
Cryptography. This is the one that should worry you, even if you never touch a quantum computer. Shor's algorithm, if run on a sufficiently large fault-tolerant quantum computer, can break RSA and elliptic-curve encryption — the mathematical locks protecting most of the internet's traffic.
The quantum computer that can do this doesn't exist yet. Current quantum computers have hundreds of physical qubits. Breaking RSA would require thousands of logical qubits, which means millions of physical qubits. We're not there.
But here's the catch: adversaries are recording encrypted traffic today with the intention of decrypting it once quantum machines can break the keys. This is called "harvest now, decrypt later." If your data needs to stay confidential for more than about ten years — medical records, state secrets, long-term financial documents — it's already at risk. The migration to post-quantum cryptography needs to start now, even though the quantum threat is years away.
What Hasn't Changed
Amid the genuine progress, a few things remain stubbornly true.
You will not have a quantum computer on your desktop. Ever. Quantum processors require temperatures colder than deep space, vacuum chambers, and precise electromagnetic shielding. They're not consumer devices. They're specialized instruments that will be accessed through cloud services, the same way you access supercomputing today.
Classical computers keep getting better. Every quantum advantage claim is measured against the best classical algorithm on the best classical hardware. That bar keeps rising. Classical researchers respond to quantum claims by tightening their algorithms, and they're very good at it.
Most computing will never touch quantum. Email, video, web servers, databases, operating systems — none of these benefit from quantum processing. Quantum computing will be a specialized tool used by specialists, not a general-purpose replacement for your laptop.
The Experiment I Actually Ran
I signed up for IBM's quantum cloud service (free tier) and ran a Bell-state experiment. The code was maybe fifteen lines of Python: allocate two qubits, apply a Hadamard gate to the first, apply a controlled-NOT, measure both.
The result was exactly what the theory predicts: the two qubits always agreed. If the first measured zero, the second measured zero. If the first measured one, the second measured one. Every time. That's entanglement in action.
The interesting part was the error rate. On IBM's simulator, the results were perfect every time. On real hardware, about 3-5% of runs produced wrong results. That's the noise floor. That's what error correction exists to tame. Running the experiment on real hardware and seeing the errors made the whole field more concrete. This isn't abstract physics — it's engineering, and the engineering is hard.
If you want to try it yourself, it takes about thirty minutes to set up. You'll learn more from running one experiment than from reading a hundred articles. I mean that literally.
The One Thing You Should Actually Do
Ignore the hype about quantum computing changing everything. Ignore the hype about quantum computing being a scam. The truth is in the middle: it's a real technology with real limitations, making real progress on real problems, on a timeline that's measured in decades, not years.
But the cryptographic deadline is real and it's now. If you're responsible for any system that stores or transmits sensitive data, you should be asking your vendors about their post-quantum cryptography roadmap. Not because quantum computers are about to break your encryption — they're not — but because the migration to quantum-resistant algorithms takes years, and the sooner you start, the less painful it will be.
NIST published post-quantum cryptography standards in 2024. Major browsers and cloud providers already support hybrid key exchange (classical plus post-quantum). The tools exist. The question is whether you're using them.
For everything else about quantum computing, watch and wait. The field is real, the progress is genuine, and the timeline is longer than the headlines suggest. If you're a software engineer, the IBM quantum cloud is free and worth an afternoon. If you're not, wait for the technology to mature and focus on the cryptographic migration — that's the part that affects you regardless of your job.
I started the week thinking quantum computing was either going to change everything or was mostly hype. I ended the week somewhere more interesting: it's a real tool that does real things, limited in ways that matter, advancing in ways that are genuine, and relevant to most people only through the lens of cryptographic security. That's less exciting than "quantum computers will solve everything," but it's true. And true is more useful than exciting when you're making actual decisions.