Double-slit experiment

The double-slit experiment is closely related to wave-particle duality: When we launch quantum objects through a small slit, they form a continuous pattern on a screen at a distance. But if we fire them through two slits, they create an interference pattern, even when launched one at a time. It's said that each quantum object travels as a wave (passing through both slits), but hits the screen as a particle. All the hits reveal the interference pattern of the quantum wave.

Wave behaviour is always present, whenever we perform single-slit or double-slit experiments. The single-slit pattern is similar to the one macroscopic objects would create, but diffraction (a purely wave phenomena) is present in both cases. It just happens that double-slit experiments can show wave or particle behaviour much more dramatically, depending on the information we acquire:

• Getting which way information: If we determine which slit each quantum object goes through, we're effectively doing a single-slit experiment each time, because we're forcing some information carriers to end at the slit's detectors, virtually closing any other open slit and preventing them from reaching the screen. We're modifying the way information would have “propagated” undisturbed, so those carriers don't fulfil the conditions to create the interference pattern. Although we think the detectors at the slits make no difference, revealing some information modifies the probability for other outcomes elsewhere (like on the Monty Hall problem). There's some time delay and re-routing involved, so we switch from instant correlations to causal relations. This way, we get the superposition of two one-slit diffraction patterns, much like the pattern macroscopic objects would create.
• Without which way information: If we only measure the interference pattern at the screen, we don't disturb the space in between trying to catch carriers “mid-flight”, so they keep delocalized the whole journey, they don't have a “checkpoint” in between. This way quantum information can keep the correlations, as if really passing through both slits at the same time, revealing the interference pattern.

QM explains the double-slit experiment using wave-particle duality and wave function collapse: The quantum object is emitted in wave/particle superposition until an observer collapses it, setting what the quantum object was the whole time (wave or particle). Then, observers discover in amazement each quantum object acted as wave or particle the whole time, in order to match the configuration of the experiment.

There are other QM interpretations, like the De Broglie-Bohm theory (or pilot-wave theory) that don't use superposition or collapse: Particle and wave coexist as different entities. But this theory is nonlocal, and it seems to go against SR. But we've seen SR allows nonlocal effects even in a universe with a speed limit.

QM's weirdness appears because we catch carriers behaving in ways that don't follow our spacetime rules. But there's no fixed spacetime background, no absolute local realism. The actual events and interactions we collect each time shape our concept of a spacetime background supporting reality. So this stability or persistence of spacetime is what we should consider “weird”.