Build a toy quantum computer at home

However, this article falls into a pet peeve of mine which is that the behavior exhibited here can also be completely explained classically -- this is also a standard demo when explaining how polarization works classically. I feel that it is worth it to at least include a footnote to that effect. The reason that I bring it up is that I (as someone who first learned classical optics, but is now learning quantum optics) personally suffered from some deep rooted misunderstandings about quantum mechanics due to having seen so many of these simplified demos which do not actually capture the quantum nature of light.

The way this article is presented it implies that one can also model quantum phenomenon using maxwells equations -- which is obviously not true. In this specific case you get the same answer, but as soon as you start looking at the individual photon statistics your answers will start to diverge. This is where the actually quantum things like Bells inequality and the Hong–Ou–Mandel effect come into play. If people had just been up front with their descriptions 'oh by the way, when you look at the aggregate behavior of photons they look perfectly classical, it is only when you look at the statistics do they behave any different' it would have saved me a lot of soul searching and misguided contempt for the quantum community.

And the idea of using sequential rotation to keep track of cumulative bias in coin flips is an interesting concept.

But ultimately I think neither one of those concepts really depends on the other in this experiment. Checking for light through polarizers is neat, but keeping track of any other rotating macro-scale object would work just as well. You can do the same thing by rotating a stick on a piece of graph paper. If it goes beyond your pre-determined test angle, you declare a bias.

As I understand it, the crazy thing about quantum computing is that you don't need to go sequentially; you can simultaneously compute every test flip in one step with qubits. That's why quantum computing could speed up certain calculations. (Note: please don't ask me to explain how.)

As I understand the experiment now, it seems like the more subtle the bias in the coin, the more times you would need to rotate the polarizer to detect the bias.

If there is something about using the polarizing filters to keep track of tries that is more efficient than using something like a stick, then I would emphasize that in your write-up.