science misconceptions

Debunking and explaining many misconceptions in science. Because there are way too many. Note: to learn anything from this, I reccomend you ACTUALLY READ THE WHOLE SECTIONS. If you don't, you may miss some very important context.

Astronomy

Not many misconceptions are here, but there are still some.

Silent spaceship booms

Actually, a spaceship exploding is one of the cases where there is sound in space. Basically all spaceships have a lot of some gas inside them, like air for breathing. When a spaceship explodes, this air is released, and thus can carry sound through space.

Gravity everywhere

Not all of space has gravity. For example, the Earth-Moon system contains a spot where gravity cancels out between the Earth and Luna. However, these spots are still very rare, constantly move around, and are unstable equilibria, so this claim is still mostly true. These spots also have bigger tidal forces than other spots, and actually they will always have spacetime curvature anyways.

General relativity

Welcome to general relativity, the SO(3,1) gauge theory, where the rank-4 Ricci tensor awaits you.

Gravity is curvature

Actually, tidal forces are curvature. Now, all gravity that we know of comes with tidal forces, but curvature, in general relativity, is defined by thing moving closer and further as measured in spacetime when moving in straight lines. A uniform gravitational field is the same as what you see under constant acceleration, and thus there is as much curvature as in flat spacetime. You're just using a different refernce frame. It's like latitude and longitude: it may seem like we have spherical geometry, and we do, but when you account for the existance of altitude, space is mostly flat around Earth.

We still use these kinds of coordinate systems since they are convinent (after all, who expresses their location in cartesian coordinates?), but for there to be curvature requires a change in gravity across space and/or time: around Earth, gravity changes to point inwards, and when a black hole forms, it changes to make a black hole. That's where the actual curvature is.

Gravity is higher dimensions

General relativity describes gravity via curvature. It is commonly taught like how a sphere has a curvature on it. However, in the case of general relativity, higher dimensions are not required. For example, on a sphere, we can describe it as a section of 3d space, as a 2d surface embedded in 3d. However, we can also imagine a 2d surface where lines are natrually weird. That's what general relativity says. General relativity uses the metric to describe spacetime. It is a tensor (basically a matrix, but we handle it more symmetrically), and it tells us how to measure distances, both lengths in space and durations in time, along paths.

The metric, importantly, lets us calculate how to make a straight line. To do this, we actually don't minimize the length (for example, the geodesic on a surface with a bump is not of minimum length). Instead, we basically make sure that for small segments of our lines, we can go to the next segment by 1. picking a direction, 2. copying our segment forward and backward in that direction, 3. extending those lines, and 4. going back to the middle. This gives us a "straightest path". Note, however, we have made no reference to anything outside of spacetime. It is like how a block of jello can bend along with the things inside it, without becoming 4-dimensional.

However, there are still some theories that use higher dimensions, since gravity is weak. For example, if spacetime has a higher dimension that loops, gravity would first diffuse into all 4 spatial dimensions, but once it has diffused fully into the looped dimension, from there on it's just inverse-square. A lot of the force would be wasted in this extra dimension, which would weaken gravity macroscopically. However, we haven't actually figured out if this is the case. For all we know, it could be a couple millimeters wide, although it probably isn't. There could also be more looped dimensions, or none. Gravity is just weird in general.

Quantum

For too long, quantum physics has not been properly repersented or used or taught, in videos and in fiction mainly. This page exists to help clear up those misconceptions, because I am absolutley tired of people saying stuff like "atoms align". Speaking of which, we'll start with atoms.

Atoms

Atoms being empty space

It is true that, if you measure the positions of the particles in an atom, it'll seem mostly empty. This is because particles are technically points. However, in practice, what we care about is the wavefunction. Let's say we have a hydrogen atom. In a "position eigenstate" picture, one where the particle's positions have been measured, and thus they look like dots, hydrogen looks something like this (ignoring quarks, and this is not and cannot be to scale):

Here, the left dot is the proton and the right is the electron. Note that the dots have to be enlarged to be visible. However, if you actually measured a hydrogen atom like this, you'd probably have blown it apart, due to the uncertainty pinciple. Most momenta the electron can have are ones where it is flying right off that proton. If we don't measure the hydrogen, it looks more like this:

This isn't exact due to SVG limitations, but you can see that the electron is smeared across the whole atom, and actually, it'd be smeared across all of space, although if I did that your browser would be extremely mad at me. Anywas, although the electron is only at a single spot at a time, its wavefunction is smeared so much that it might as well cover the atom. In fact, it's pretty helpful (for me) to think of how the wavefunction moves instead of how a single dot moves when it comes to quantum mechanics (in that interpretation, the wavefunction here is holding itself up from the proton's). You can also see that the proton is fuzzy, or, you could if you happened to zoom in on the center a lot more than usual (trust me it's there, you can check the html and the svg is there). It's 1,836 times smaller than the electron cloud, and so unless you're doing nuclear physics, this is where you can stop doing quantum stuff (usually). However, clearly, this atom is not empty. The electron cloud is filling it up. It also wouldn't actually make sense for 2 dots to be "empty" or "not empty" anyways, as they are literally dots, not to mention the fact that quantum weirdness makes you question what even is empty (the electromagnetic field is filling the atom too).

Atoms not touching

It is said that atoms simply cannot touch, and that you also cannot touch things. This is false. Ok, for one, "touching" basically always is defined macroscopically anyways. Even if we use quantum physics, however, you can still touch stuff. For example, electricity: electricity needs things to touch for it to pass, as the electron clouds of parts of metal must overlap enough so that electrons can go between them. Already, this is clearly touching if the electron clouds must overlap, and indeed they can: how else do wall plugs and the internet work? And yes, you can touch things this way too, if you couldn't, electrocution wouldn't be a problem. Even ignoring this, atoms still fully share electrons in covalent bonds, and even with no electrons, the neuclei of atoms can join together (ex. the sun), which is beyond "touching".

Now, yes, it is true that things like pauli exclusion and electrostatic repulsion are the only things stopping you from going into your chair, and that the neuclei of atoms don't toich in everyday life (hopefully), however if those didn't exist, your atoms would be fusing with your couch right now. Actually, you wouldn't have a couch. And probably no life as we know it. So yeah, even with quantum physics, yes, atoms touch, they do it all the time, and you can touch atoms yourself too.

Aligning the atoms

No, that is NOT how quantum tunneling works. Ok, let's use a macroscopic analogy. Get (or imagine) a transparent bag, like a ziploc. The photons (light) passing through it is like you trying to pass your hand through a very very very very thin wall. Actually, more photons are let through than hands would be through a wall (graphene, which is a single atom thick, can be pushed around), but let's just ignore that for now and say that your hand is also very thin. Now, fold this bag. You previously has 2 layers stopping photons. Now you have 4. Keep doing this. Eventually, it'll look white. Why? Well, if a photon is to pass through, it must get through all layers. Even though each layer only stops some of the photons, together they stop basically all photons.

Now, a normal wall may have around, let's say, a billion atoms. It probably has more. If there is a 99% chance your hand passes through each layer of atoms, the chance of you passing through would be so small that even Wolfram|Alpha, the best calculator I can think of, thinks it's exactly zero. This is why you don't pass through walls.

Now, even if you just wanted to align the atoms and din't need quantum tunneling, most materials would be really bad for this. For atoms to be aligned in the first place, you'd need them in a crystal structure. For example, a grain of salt has a cubic crystal structure, just like a minecraft world. However, usually, there will be many "grains". Each grain is that crystal structure, but turned a bit in some way. These would make it impossible to align these grains of salt properly. Even if you did, the electron clouds smear out the bonded atoms anyways, and any increase in chances of quantum tunneling would still be extremley low.

Note: in the very very very very very rare case you quantum tunnel through a wall, you're probably going to be stuck for a while if your hand only tunnels partially.

Measurement

Not only is measurement not required for quantum physics to work out due to decoherence, but it also leads to way too many misconceptions. Maybe it is better to just use many-worlds.

Looking at 2 slits

There is an experiment called the double slit experiment. You probably already know this, but if you don't, basically you cut 2 very small holes in something like paper, spaced only a couple micrometers apart, and shine a laser through it. The setup's common display already has a misconception. Mainly, people commonly draw the slits pretty big. Now, this does help for visualization, but it helps a lot if you display a massive bannana for scale (or something like that). After all, far-away slits would barley do anything other than stop the laser. Anyways, this laser is passed through, and you get an inferference pattern. This can be done with water waves too for intuition, where the high and low parts cancel out and make some regions not move up and down.

Now, here comes the biggest problem: the way people show the measurment. In a real setup, if you put a polarizing filter that changed the polarization of light (ex. a glass of sugar-water that is filled just enough so polarization turns 90°) over one slit, this destroys the interference. It can be though of as measurment, but it's more helpful (at least for me) to think of it as the following: previously, the light was able to cancel out itself because it is pointing in the same direction, but with one path turned 90°, all it does is move left/right, which cannot cancel out anything. Either way, yes, that would happen.

Now, here's the problem: many depictions show people placing a camera (or something like that) to the side of the slits. You see the problem?

Well, since we're using laser light, this will not actually affect or measure the light at all, because THE LIGHT ISN'T GOING THERE.

Seriously, the whole main thing about a laser is the straight beam. Have people forgotton that??? Anyways, really, the reason that the measurment destroys interference (when done properly) is that what we are doing make sit so there is something different, other than location, about the 2 paths. Remember how the "tilted" (polarized) light couldn't interfere with light polarized the other way? Here, what happens (usually) is that one slit, when light goes through it, either does the polarizing thing or ends up making the light have an effect on something else, the latter of which requires it to be absorbed, and usually when the photon is emmited back, its phase (the part of its oscillation it is on, which constantly increases but loops around) is scrambled, and also sometimes its polarization, which destroys interference. Note that this happens to each photon seperatley, in fact, the only thing forcing quantum mechanics onto the double slit experiment is the fact that it works with single photons.

Retrocausality

There is a variation on the double slit experiment. We take a laser, as before, but then use a special crystal (a BBO crystal), which takes a high energy photon and halves its frequency into 2 lower energy photons, and also doesn't destroy quantum effects (somehow this actually works). 2 of these crystals are put on both paths. One path on each crystal leads back in the slit experiment, while the other goes on to some beam splitters, which decide weather or not we have measured the photon. We then filter results from weather or not we measured the photon or not, and yay, we can change weather or not we see interference.

I have no idea how people got the idea of "quantum mechanics gives retrocausality".

Really, the only reason why the filter determines weather or not we see an interference pattern is because we're filtering on weather or not we measured the photon. It's that simple. In fact, I can think of a classical analouge: have a coin that is on heads, flip another coin, if that other coin is on heads we flip the first, and now ee filter the first coin's result by the 2nd's result. Now, by choosing the filter, according to this retrocausality logic, we can control weather or not the first coin can come up tails. No, but seriously, where did people get retrocausality from this?

Conciousness causing measurment

No, conciousness does not cause measurment. If it did, philosophers would've easily figured out the philosophical zombie paradox. There is also nothing in our brains to cause quantum collapse. Even if it could be, it would make no sense for us to evolve such a thing. End of story.

Quantum invincibillity

No, quantum immortality would not make you immune. For one, please do not test it out, as you would be risking your life on something that doesn't even make physical sense. Anyways, quantum immortality only would garuntee that you don't stop perceiving things, and so if an anvil falls on you, it is very well possible that instead of quantum tunneling fully through you and saving you, it instead quantum tunnels partially through, but still gives you major scratches, or worse. Although the whole idea of quantum immortality itself kinda doesn't make too much sense, as conciousness doesn't even seem to be based in the physical world (as there is no law of physics providing a way for beings to experience this universe). Also, quantum mechanics is supposed to be a statistical theory for how particles move, and is not supposed to have philosophy out of all fields in it.

Wave-particle duality

Wave-particle duality exists in a way, but not as it is usually said. Measuring a particle, even under the copenhegan interpretation, does not actually collapse a wave to a point. If it did, you would've used infinite energy and made a black hole, which would trap all that quantum information and then blip out of spacetime as we know it, maybe making another universe or whatever. Instead, it just makes the wave way smaller. Meanwhile, in quantum field theory, wave-particle duality basically says that once you measure the amount of particles, there are some amounts of wave that you simply cannot have, since they'd require half-particles. Also, the many-worlds interpretation basically just keeps everything as a wave, which while very useful for sidestepping the measurment problem, can get very complicated due to entanglement.

Quantum field theory

Welcome to quantum field theory, where the fields can have positive curl and negative curl at the same time, while you GPU slowly melts itself with quantum harmonic oscillators.

Zero-point energy

Despite what popular science, GTNH, and GT:CEu modern would have you thinking, you CANNOT get energy from the vaccum. Zero-point is defined as the lowest energy. Luckily for us, in quantum field theory for a free quantum field, zero-point energy exists for the same reason it exists in a quantum harmonic oscillator. This makes things thousands of times simpler. Let's at the QHO's ground state:

The line in the middle is zero (classically the minimum energy). We can see that the wavefunction extends outwards. Now, imaging making this bigger. More of the wavefunction would be further, meaning more energy. Ok, makes sense. But now, imagine making the wavefunction smaller. This increases uncertainty in momentum, and thus increases kenetic energy, making it require energy to zero the wavefunction! Same thing with zero-point energy: trying to take from it is actually adding energy to the vaccum. Of course, the harmonic oscillators in quantum field theory are different (they appear in fourier space), but the idea does still hold very well. In fact, the quantized energy levels of these oscillators are what gives rise to the fact that you can't have half an electron. Neat!

Quantum computing

Is it hype? Is it not? Noone really knows, but it's probably not that useful.

Internet destruction

No, the internet will not be destroyed. For one, a fair amount of sites don't use RSA. Wikipedia, for example, uses elliptic curve encryption, which uses clever geometry and somehow manages to be quantum-resistant. For websites that do use RSA, like this one, we haven't actually found a good way to implement Shor's algorithm anyways. This is because toffoli gates, which flip a qubit if 2 other qubits are both on (while perserving quantumness), are very hard, but a lot are needed for shor's algorithm in quantum. It must also be noted that quantum computing, for now, still requires temperatures at and below 1 kelvin, which would be a lot for some bad actor to get just to crack some RSA keys. So yeah, we're fine.

Quantum parallelism

This one seemingly stems from the many-worlds interpetation. I'm not sure. Anyways, yes, quantum computers do act on a lot of states at once, that's accurate. However, this does not garuntee gpu-like performance. Why? Well, most of this information is encoded in a quantum state. However, when we check on our qubits, we only get 1 possibility out of the many we could get. This means to get useful stuff, we need to be very clever and "remerge" these quantum states back into one, so that we can take fewer measurements: after all, each measurment requires us to run our whole algorithm, whatever we do.

Now, this filtering is still, sometimes, possible, although there's another key thing: phase will be missing no matter what. Phase helps determine how interference will happen, but measurment scrambles it entirley (unless you're dealing with things like radio waves where phase changes very slowly). Finally, even if quantum computing was parallel, we already have GPUs, which have the advantage of being able to copy-paste properly, so it's not that useful (the useful thing is actually interference).

Quantum is useless

You might get the impression that quantum computing is fully useless. However, it does actually have a fair amount of uses. For example, quantum field theory requires you to deal with a superposition of states a field can be in. There are a massive number of states a field can be in, compared to those a particle can be in. Techniques like mode decomposition do help, however those are mainly for free fields, ones without interaction. Fort things like quantum electrodynamics, it's better to use a quantum computer: we can store the field simmilarly to on a classical computer, but it can natrually be in a superposition of states, since it's a quantum computer. Even better, these states will interfere and do stuff like that properly, as a pure consequence of quantumness.

Another thing quantum would be useful in is chemistry. For example, quantum simulations may help us understand nitrogen-fixing, which would help us make better fertilizer. Now, most things that classical computers do will probably not need quantum computing, but it could be possible to add a quantum processor to a computer for the cases where quantum is useful.

Entanglment is FTL

Entanglment does not allow you to send information faster-than-light. It may seem that way, but it's more of you inferring a fact about something faw away, not information transfer. Take, as an example, 2 coins. We flip the first coin while it is connected with a straw the the 2nd so that their heads/tails state is always the same. Now, we break the straw and make sure our coins do not flip themselves ever, and bring them on a journey far, far away. Now, if I look in the first box, I will immediatley know that the 2nd coin is whatever the 1st coin shows. The coins did not transfer any information. Same with quantum physics. We simply infer the state.

Now, let's say we have an entangled pair. Let's use 2 qubits, q0 and q1, both set off. We "randomize" q0 with a hadamrd gate, which puts it in a superposition of on and off. We then use a cnot gate, with q0 ad the control and q1 as a target, which entangles q0 and q1 such that they are the same state. We yeet them far away, like with the coins. Now, surley, if we measure q0's on/off state, then it must be the same as q1's on/off state, right? Yes.

Now here's the part you might think breaks physics: what if we hadamard q1, measure, and then hadamard again? When we hadamard-measure-hadamard, we basically measured a quantity "orthagonal" to on/off, the hadamard states +/-. This means our measurment on q0 will not be dependent on q1's measurment! However, this does NOT let us transfer information. q0 is still in a superposition of on/off, and will still act the same when we measure it.

To have any hope of knowing which way q1 was measured, we must transmit that information, either through this setup again or by classical means. Eventually, we'll need to use classical communication, which we know is slower-than-light. In the coins analogy, what we've done is simply re-flipped the 2nd coin, and expected the 1st coin's state to measure different. Of course no information is sent faster-than-light, since entanglment is literally just a correlation. So no, quantum does not mean FTL information transfer. Its FTL features are as FTL as our classical "entangled coins" analogy, and that's it.