What you're about to see
Fire light at a wall with two narrow slits cut in it. Put a screen behind the wall. What do you see on the screen?
Classical intuition says: two bright lines, one for each slit. The light goes through one or the other and lands behind. Simple.
What actually happens is so strange that physicists are still arguing about what it means, a hundred years after the first measurements.
Fire photons
Try this: Start on 2 slits · unobserved. Hit Start stream. Watch single photons hit apparently random spots — but the pattern that emerges tracks the cyan curve, not the gray classical one. Then switch to 2 slits · observed. The moment you peek at which slit each photon went through, the interference vanishes and the cyan curve snaps onto the classical one. The act of measurement physically changes the experiment.
What you should notice
Hit "Two slits" and click Stream. Each individual dot lands in an apparently random spot. But run a few hundred and a pattern emerges: bands of bright stripes alternating with dark gaps. That's an interference pattern — the kind of thing you see when two water waves overlap, with crests adding to crests and crests canceling troughs.
This is the first weird thing. You're firing single photons, one at a time. Each one is a particle — when it hits the detector it leaves a single dot, not a smear. But the statistics of where they land matches a wave interference pattern. Each photon is somehow interfering with itself.
This rules out a lot of intuitive explanations. The photons aren't bumping into each other — they're fired one at a time. They aren't literal water waves — each one leaves a single point of impact. They're something else.
The second weird thing — measurement
Now switch to "Two slits, observed." This adds a detector at each slit that records which slit each photon passed through. Hit Reset, then Stream again.
The interference pattern vanishes. You get two bumps, one behind each slit. The classical picture. The pattern that "ought" to be there if photons are particles.
Nothing about the physical setup has changed except whether information about which-slit-each-photon-took is recorded somewhere. That information's mere existence — whether or not anyone looks at it — destroys the interference. You can run this experiment yourself at home with a laser pointer and razor blades.
What does this mean?
There are several interpretations of what's actually happening, and physicists still don't agree:
- Copenhagen interpretation (the textbook view) — before measurement, the photon doesn't have a definite path. It's described by a wave function that goes through both slits simultaneously. Measurement "collapses" the wave function to a single outcome. Asking which slit it "really" went through before measurement is a meaningless question.
- Many-worlds interpretation — the photon goes through both slits, and the universe splits into branches where it went through each. Branches interfere when they recombine on the detector. When you measure, your branch entangles with a specific photon-path branch and you stop seeing the other one.
- Pilot wave / Bohmian mechanics — the photon really does take one path, but it's guided by a "pilot wave" that goes through both slits. Measurement disrupts the pilot wave. Less popular today but logically consistent.
All three predict the same experimental results. The disagreement is purely about interpretation — what's "really" happening before you measure.
Why this is the foundation of quantum computing
Quantum computing exploits exactly this property: that quantum particles can be in superposition — taking multiple paths at once, or being in multiple states at once — and that those parallel possibilities can interfere constructively or destructively. Every quantum algorithm is, fundamentally, an interference engine: the circuit is designed so that paths leading to wrong answers cancel out and paths leading to right answers reinforce.
A qubit in superposition is doing the double-slit experiment, but with abstract logical states instead of physical paths. The wave-function math is identical. The Bloch sphere is just a clean way to draw what would otherwise look like an interference diagram.
Things you can verify
- Single-photon interference is real. Done first in 1909 (Taylor) with dim light, and definitively with single photons in 1989 (Tonomura).
- Molecules do it too. Anton Zeilinger's group showed buckyball (C₆₀) and even larger molecules (up to 2,000-atom proteins) producing interference patterns. Quantum behavior is not limited to small particles — it's a property of how nature works.
- The "delayed choice" version is even weirder. John Wheeler proposed measuring which slit after the photon has already passed through. Done experimentally — the result still depends on the measurement choice, even though that choice was made after the photon allegedly "committed" to a path.
Want to go deeper?
Now that you've seen the double-slit pattern, the rest of quantum mechanics has a place to live in your head. The two best next steps:
- The primer — 16 concepts from qubits to fault-tolerant computing. Includes the interactive Bloch sphere, entanglement explorer, circuit builder, Grover's algorithm walkthrough.
- The glossary — every quantum term defined.
- Best books, videos, podcasts — including Quantum Country by Andy Matuschak & Michael Nielsen, the gold-standard interactive book on this topic.