The narrative of quantum computing is a magic show. First, the rabbit appears: a problem unsolvable by classical computers. Then, the hat appears: the quantum machine. But as research progresses, the rabbit tends to disappear.
The industry relies on the concept of Quantum Advantage—the threshold where a quantum computer outperforms a classical one. But "outperforming" is a moving target. Classical algorithms are not standing still.
Click the proposed use cases to update their status based on recent findings.
When we look closely at specific examples, the story collapses. We were told only a quantum computer could simulate complex molecules. Then, classical teams optimize their code and solve it. The "impossible" problem was just a hard problem we hadn't optimized yet.
Marketing teams love the Traveling Salesman Problem (TSP). A salesman must visit $N$ cities via the shortest route. It is NP-Hard. The assumption is that quantum computers, using superposition, can check all routes at once.
This is a misunderstanding. Quantum computers manipulate probabilities, they don't simply "try everything." For optimization, classical heuristics (guessing intelligently) are remarkably efficient.
Notice the difference? The classical approach (Nearest Neighbor) isn't perfect, but it's instant. The "Quantum" approach (often conflated with brute force in pop-sci) faces an exploding search space ($N!$). Even with Grover's Algorithm providing a square-root speedup, the overhead and noise make it non-competitive against classical approximations.
There is a problem no one talks about: Thermodynamics.
To make quantum computers work, we need Error Correction. To get one perfect "Logical Qubit," we need thousands of noisy "Physical Qubits" to correct it. All of those physical qubits need to be cooled to near absolute zero.
If we need millions of physical qubits to do useful chemistry, we aren't talking about a chip in a laptop. We are talking about a facility consuming the power of a small city, just to run the error correction cycles.
We are often told this is the "Transistor Moment" for quantum tech. This analogy is flawed.
The Transistor: Became smaller, faster, and cooler over time.
The Qubit: Requires massive macroscopic equipment (dilution refrigerators, lasers, vacuum chambers) that does not scale down linearly.
The skeptical view isn't that quantum computing is impossible physics. It's that it is impractical engineering. If the error correction requires more energy than a supercomputer would take to just solve the problem classically, where is the advantage?