A quantum computer is not a faster version of your laptop. It is a different machine that plays by different rules, and it is very good at a very short list of jobs.
For everything else — your email, your photos, your games — an ordinary computer wins, and always will.
It only ever does three things
It sets things up. A row of qubits all start in the same known state.
It stirs. This is the clever part, and we will spend most of the course here.
It looks. You read the qubits, and out come ordinary 0s and 1s. That is the only kind of answer you ever get.
The line everyone gets wrong
You have probably heard that a quantum computer "tries every answer at once".
It does hold every possibility at the same time. But you only get to look once, and looking gives you a single answer. If the stirring in step two was lazy, that answer is just a random guess.
What it is genuinely good at
Molecules
Working out how atoms behave — new medicines, better batteries
Codebreaking
Cracking the math that protects most of the internet today
Searching
Finding a needle in a haystack roughly a square-root faster
Notice what is not on that list: general planning problems, logistics, and most of what people mean by "AI". Those show up in adverts far more often than in results.
Where things actually stand
Today's machines are small and jittery. Qubits forget what they were doing after a fraction of a second, and roughly one operation in a few hundred goes wrong.
That is improving. In 2024, Google's Willow chip showed error-fixing that gets better as you add more qubits — the first time that had been demonstrated. It is a genuine milestone, and there is still a long way to go.
Worth remembering
- A quantum computer sets up, stirs, then looks — and looking gives plain 0s and 1s.
- The magic is in making wrong answers cancel out, not in trying everything at once.
- Molecules, codebreaking and searching are the real uses. Most other claims are noise.
- Machines today are small and error-prone, but error-fixing has started to work.