In 1935 Einstein came up with a thought experiment that showed quantum mechanics breaks one of the most sacred principles in physics, that nothing can go faster than the speed of light.
Physicists assumed he was wrong.
They thought that at 56, Einstein was an old man, past his prime, and just unable to accept the new theory of physics because it was too radical.
But 30 years later, one man stumbled across Einstein's forgotten paper when he realized something.
The prediction could actually be tested.
When scientists ran the experiment, they were shocked.
Quantum physics really does break the universal speed limit.
We're obliged to invoke something like actions going faster than light from one place to another.
This is a video about one of the spookiest and most misunderstood experiments in all of physics, and it might even be the strongest evidence we have that we live in many worlds.
If the sun were to disappear all of a sudden, how long would it take until we noticed and were released out into space?
Newton's theory says that gravity acts instantly across any distance.
So if there's a change in gravity, we should feel it immediately.
But Newton himself was disturbed by this.
That one body may act upon another at a distance is to me so great an absurdity that I believe no man who has a competent faculty of thinking can ever fall into it.
But in 1905 Einstein realized action at a distance isn't just absurd, it leads to outright paradoxes.
Einstein had discovered that observers moving at different speeds can disagree about when events happened.
Let's say you see two things happen at the same time.
An observer speeding past would see it differently.
To them, one of these happened first, and both points of view are equally valid.
But in the case of gravity, this leads to disaster.
Say you see the Sun disappearing and Earth flying off at the same time, as Newton predicted.
Then the other observer sees something impossible.
They see the Earth flying off first, even while the Sun is still there.
And it should, of course, be pulling the Earth in.
So to them, it looks like cause and effect are reversed.
The only way out of this paradox is to reject the assumption we started with.
So gravity can't be instant.
It took Einstein 10 years to fix this issue, and in the process he completely overhauled our understanding of gravity.
Gravity is caused by the bending of space-time.
When there's a change in gravity that only affects the local space-time and then that ripple spreads out to nearby regions, which spread farther out until eventually they reach us.
This theory of gravity is local, because effects spread from place to place at the speed of light instead of being instant.
From our frame of reference, if the sun disappeared, that ripple would take about eight minutes to reach us.
Another observer might disagree about the length of the delay, but now we all agree that the sun disappeared first.
This is why nothing can go faster than light.
The delay between cause and effect ensures all observers agree on the order.
After Einstein fixed gravity, all of classical physics obeyed this important rule.
But then, Einstein studied the new theory of quantum mechanics and made a terrible discovery.
This is one of the most famous photographs in physics.
It was taken at the 1927 Solvay Conference, where the architects of the brand new quantum theory gathered to discuss it.
Around 60 of the attendees would win Nobel prizes, but Einstein thought they'd gotten something fundamentally wrong, and this was his chance to prove it.
So he took to the stage with a thought experiment.
Imagine you fire a single electron through a narrow slit toward a circular detection screen.
Well, quantum mechanics says that this electron has some sort of wave associated with it, called a wave function, which spreads out through space as it travels.
When the electron hits the screen, you detect it at a single point.
Where it turns up depends on the amplitude of the wave.
If the wave is very big in a particular area, the electron is more likely to turn up there.
Let's say it appears here.
So far, everyone was following.
This is what quantum mechanics predicts.
But Einstein's next question surprised them.
Why doesn't the electron turn up at this other spot a moment later?
There's only one electron, so we can't detect it twice.
But the way quantum mechanics ensures this is that when the electron was detected at the first spot, its wave function collapsed to zero everywhere else instantly.
That's why the probability of finding it at the second spot is now zero.
There's no longer any wave there.
But Einstein asked the audience to think about what this means.
The measurement here must instantly affect the wave function over here, no matter how far apart these locations are.
In other words, quantum mechanics requires instant influences across distance.
It violates locality.
Einstein concluded his talk by saying this is an entirely peculiar mechanism of action at a distance and that this implies, to my mind, a contradiction with the postulate of relativity.
Einstein's argument was so simple and his talk so short that people didn't know what to make of it.
One audience member said I feel myself in a very difficult position because I don't understand what precisely is the point which Einstein wants to make.
No doubt it is my fault.
That man was Niels Bohr, the most influential figure in quantum physics at the time.
Bohr's institute in Copenhagen had become the hub for the new field.
Dozens of young scientists like Werner Heisenberg came to learn from him.
As one of his disciples remembers, Bohr had invited a number of us to his home, where we sat close to him, some literally at his feet on the floor, so as not to miss a word.
Bohr wasn't the one who wrote the mathematical rules of quantum mechanics.
Instead, he told everyone what they meant.
While others were confused by the theory, Bohr offered answers.
His philosophy became known as the Copenhagen interpretation of quantum mechanics.
My general understanding of the Copenhagen interpretation is you have the wave function.
It describes everything that you can know about a particle or a system and it evolves according to the Schrodinger equation and at some point you're going to make a measurement and at that point the wave function collapses.
I think that one bit of that that you said was like the wave function is all you can know about the particle, and I think that was like a pretty important point to bore.
As Bohr would put it, it's wrong to think that the task of physics is to find out how nature is.
The job of physics is just to predict measurements in the lab, which quantum mechanics does incredibly well.
As for what the electron is doing when you're not looking?
Well, to Bohr, that question didn't even make sense to ask.
The wave function tells you everything physics can or should tell you.
Einstein couldn't stand the Copenhagen Interpretation.
In a letter to his ally Schrödinger, he called it a tranquilizing philosophy, or religion.
Einstein felt his thought experiment exposed a critical weakness in the Copenhagen interpretation.
He'd shown that the way the wave function collapses is non-local, and so he reasoned maybe the wave function is the problem.
Maybe it's not the best way to describe the electron after all.
He may not have convinced Bohr of this during his talk, but he was determined to do it during the rest of the conference.
Physicists tell a version of this story.
You know that you will find in physics textbooks and in pop science books.
And that you know.
Physicists tell amongst ourselves that what happened was Einstein and Bohr had a great debate And Einstein was unhappy with quantum mechanics because it was fundamentally probabilistic.
He tried to show that there were perceivable experiments that you could use to get around those uncertainty relations.
And Bohr showed over and over and over again that you couldn't do that.
And eventually everybody agreed with Bohr.
That's Adam Becker, author of What is Real, a great book about the history of quantum mechanics.
As he explained to us.
Bohr may have just misunderstood the purpose of Einstein's thought experiments.
We have documented evidence of this in at least one case.
Einstein described a thought experiment that involved a box of photons and a mirror.
Its purpose was to show the non-locality of the Copenhagen Interpretation in action.
Bohr just misunderstood it.
And when he recounted it to others later on, he drew a little diagram of what Einstein's thought experiment setup was.
It just didn't have the mirror in it at all.
And yet this is taken as like the great victory for Bohr over Einstein, which is crazy.
History is written by the victors, right?
To understand what Einstein was arguing for, think of the relationship between Newton's gravity and general relativity.
Newton's theory works well in most situations, but in that theory gravity is a non-local force, leading to paradoxes.
This was the motivation for coming up with Einstein's general relativity, which is local.
Einstein believed the same logic applied to quantum mechanics.
His thought experiment revealed that quantum theory is non-local.
So, just like with Newton's gravity, quantum mechanics must not be the final theory.
There must be a local one that will replace it.
And as a bonus, he thought this new theory might even unify gravity with the quantum world.
It would be hard to imagine coming to the final theory right away.
And yeah, and the fact that you can see paradoxes like this would make you think there's gotta be more to it that we just don't have yet.
But Einstein hadn't even persuaded Bohr that quantum mechanics really is non-local.
So in 1935 he made one last attempt to convince the community that there was a contradiction between quantum mechanics and relativity.
With the help of two younger colleagues, Boris Podolsky and Nathan Rosen, he formulated another, even more striking thought experiment that shows the non-locality of quantum mechanics.
This paper is now known as the EPR paper, after its authors.
Here is a simplified version of their thought experiment.
Imagine a single high energy photon suddenly becomes two particles.
One of them is an electron, and to conserve total charge, the other is a positron.
Since one is negative and the other is positive, they cancel out.
But both electrons and positrons have a property called spin and, like electric charge, this also needs to be conserved.
If the light started out with zero spin, well then the two particles together must have zero total spin as well.
For example, if the direction of the electron spin is this, the positron has to have spin in the opposite direction so that they perfectly cancel out.
But the electron spin could have been this instead, or this.
All of these possibilities are valid, so the rules of quantum mechanics say that the electron does all of these possible things at once until it's measured.
It's not just that we don't know what the spin is, the electron really is doing everything.
The only restriction is, whatever the electron is doing, the positron must do the exact opposite.
This also means that when the electron is measured and its state is determined, so is the positron's.
This is what we mean by entanglement.
The two particle states depend on each other.
But how do we measure the particles and force them to do one thing?
Well, for that, we use the Stern-Gerlach machine.
It's essentially a strangely shaped magnet, and it's how we measure spin.
The orientation of the magnets determines what axis you're measuring the spin in.
For example, if the machine is like this and we shoot in a particle with spin in the positive Z direction, it will certainly go to this spot.
We'll call plus.
If instead, a particle has negative Z spin, it will certainly go down to minus.
So this Stern-Gerlach machine measures spin in the Z axis.
So what happens when we put in one of our entangled particles?
When the electron goes into this machine, it either goes to plus or to minus with 50-50 probability.
Let's say our electron goes to plus.
Well, this means it went from being in an indeterminate state to positive z-spin.
But what about the positron?
Well, the only way to conserve spin is if it's now in the negative Z spin state.
When it's measured, there is a 100% chance it's minus.
It has to be that way to conserve spin.
But the authors of the paper realized there's something very odd about this result.
To see what's wrong with this, let's imagine that the electron and the positron carry these envelopes with them.
These envelopes represent the state of the two particles.
Until they're measured, both of the particles are in a superposition of being plus and minus at the same time.
So both options are in the envelope.
But now let's move the positron to someone who's far, far away.
In this analogy, opening the envelope is like measuring the spin of the electron, but that causes the wave function of the electron to collapse to just one possibility.
In this case, it's plus.
But what happens to the other envelope far away?
Well, it needs to instantly collapse to minus, because otherwise, when the experimenter opens their envelope, they have a chance of seeing plus, which would violate the conservation of spin.
But if it needs to collapse instantly when the electron is measured, then how does it know what to collapse to?
It must receive intel from the far away electron, but that message has to travel much faster than the speed of light to get to the positron in time.
And so, with this argument Einstein, Podolsky and Rosen had shown that the Copenhagen interpretation of quantum mechanics really is non-local.
Einstein had already shown this in his conference talk, but this argument was even more decisive.
It does seem like it's the same thing, but now it's ramped up and you've got these two separate particles to do those two separate measurements and one measurement influencing the other measurement definitely feels wrong.
Yeah, exactly.
I think he really realized that it's measurements that are the problem in quantum mechanics.
The wave function of a single particle or of this pair of particles can end up spread over vast distances.
That isn't itself an issue.
But when the wave function collapses, the information about that collapse needs to spread everywhere.
The wave function is.
That's what makes quantum mechanics non-local.
The EPR paper didn't just point out this non-locality issue.
They proved that there is only one local alternative theory for explaining this experiment.
In this local story.
Instead of the electron choosing whether to be plus or minus when it's measured, it actually makes that choice when it's still in contact with the positron.
There's some random way that this plus and minus gets put into these two envelopes is why the plus and minus are called hidden variables, and because this alternative theory assigns these hidden variables in a local way, while they're still in contact with each other, rather than over a big distance, we call this a local hidden variable theory.
Now this local hidden variable theory is going to be able to explain this experiment really simply.
Let's pass away the positron, And now, when the electron is measured as a plus, it doesn't have to rush to tell the positron.
The positron already knows.
There is no action at a distance.
This local hidden variable story is so much more sensible than the quantum one.
So we're forced to accept one of two explanations for this experiment either nonlocality, like the Copenhagen interpretation of quantum mechanics, or a local hidden variable theory.
Given that nonlocal action at a distance contradicts relativity, Einstein thought this was definitive proof that the Copenhagen interpretation of quantum mechanics is wrong and therefore there must be some local hidden variable theory that will replace it.
Einstein showed us that quantum mechanics allows influences that seem to travel faster than light.
But with today's video sponsor, you can have your own spooky action at a distance.
With one click, NordVPN can teleport your digital presence anywhere in the world.
It's not actually faster than light, but it might feel like it.
NordVPN has thousands of servers in over 120 countries, which means you can access content that would otherwise be region locked.
This way, you can access all your favorite content from wherever you are with virtually no lag.
But it's not just about access.
Every time you connect to public Wi-Fi at a coffee shop, an airport or a hotel, you're potentially exposing your data to anyone on that network.
NordVPN encrypts your connection, so even if someone intercepts your traffic, all they see is gibberish.
And with their strict no-logs policy, your browsing stays private.
One account covers up to 10 devices, so you can protect your phone laptop tablet, even your router.
When you use myLink, you'll get a huge discount on a two-year plan plus four additional bonus months.
Get it at nordvpn.com slash veritasium.
It's risk-free with Nord's 30-day money-back guarantee.
So go check it out at nordvpn.com slash veritasium.
You can scan this QR code or click that link down in the description.
I want to thank NordVPN for sponsoring this part of the video.
And now back to Bell's theorem.
The EPR paper certainly got a lot of attention.
Without asking Einstein, Podolsky leaked the paper to the press and the story ended up in the New York Times.
The prodigious harvest of the day's intelligence is reached.
Einstein was the most famous scientist in the world and he was going after the strange but successful theory of quantum mechanics, so of course the press loved it.
But what did scientists think of the argument itself?
So the reaction of the physics community was at first mixed.
You know, there were some people, sort of old allies of Einstein's, who were very happy with it.
Schrödinger being sort of at the top of that list.
And in fact in an attempt to clear up some of the misunderstandings that people were having about the EPR paper, Schrodinger publishes the thought experiment known as Schrodinger's cat.
Sort of back up Einstein and show the kind of problem that he and Einstein had with quantum physics.
Or meanwhile, it's like, oh my God, what the hell is this?
He must be wrong.
How do we show that he's wrong?
Then Bohr, ultimately, you know, in his sort of painful and complicated style, ends up coming up with a response to the EPR paper.
This response is sort of famously obscure and difficult to understand.
And I have read it in detail.
And I will tell you, Bohr's reply is either nonsensical or makes some actual mistakes.
Well-turned sentence which I believe Bohr took a great deal of trouble in formulating and whose meaning is just absolutely obscure to me.
Bohr said in his reply to EPR, in his multiple replies to EPR, that there is no question of anything, though
So the ultimate reaction of the physics community, at least in the immediate years and decades following the publication of EPR and Bohr's reply in 1935, was to think that Bohr had settled it with his reply, even though people didn't really understand what Bohr had said.
Two decades later, Einstein died, still questioning quantum mechanics.
But the majority of the physics community had moved on without him.
Bohr, however, never forgot about the EPR paper.
In 1962, seven years after Einstein's death, Bohr gave an interview about Einstein and he lamented that Einstein wasted decades on fruitless thought experiments because he simply could not accept quantum mechanics.
It was terrible that Einstein fell in that trap to work with Podolsky, Bohr said.
Rosen is worse from my point of view.
Rosen even today believes the EPR thought experiment.
Podolsky has given it up, as far as I know.
The whole idea is absolutely nothing when one really gets into it.
You may think that I say it too strongly, but it is true.
There's absolutely no problem in it.
The next day, Bohr took a nap after lunch and never woke up.
And so, after many decades, the Einstein-Bohr debate was over.
Bohr's authority was part of the reason the EPR paper didn't get the attention it deserved.
But there was another reason physicists ignored it.
In the EPR experiment, both theories Copenhagen quantum mechanics and Einstein's local hidden variable alternative make exactly the same prediction.
You get the same results either way.
Debating two different interpretations of the same experimental result seemed like armchair philosophy, not real physics.
The Copenhagen interpretation makes good predictions, so why not just teach that and move on?
It just seems like you know, shut up and calculate.
I think is the message that kind of gets pushed.
The general attitude was this is done.
Who cares?
None of this matters.
It's all settled.
Einstein and Bohr had a big debate about it and Bohr won.
Do you think you're smarter than Niels Bohr?
Do you think you're smarter than Albert Einstein?
It seemed like it would be impossible to resolve this debate until another physicist turned his attention to it.
John Bell was an undergraduate student shortly after World War II, in this new era of physics, and so, of course, he was taught the Copenhagen Interpretation.
John Bell's doubts about quantum mechanics, by his own recollection, showed up basically the minute he learned it.
In his first quantum mechanics class he was, you know, getting pretty upset with the instructors and saying you're being too vague.
What the heck do you mean about measurement?
Bell was never fully satisfied by the answers he got about the foundations of quantum mechanics.
But when he was doing his PhD, he was encouraged to study something a little bit more respectable.
And so he studied nuclear physics and went on to have a very accomplished career at CERN.
But after eight years of working in particle physics, in 1963, he took an academic sabbatical.
And finally, he had time to focus on his doubts about quantum mechanics.
He said, I always knew that it was waiting for me.
He began by reexamining the old debates and the papers that most physicists had long since dismissed as philosophical distractions, including the EPR paper.
After this research, he said, The question was, could you prove which one using an experiment?
An experiment of this sort seemed much more feasible by Bell's time.
The EPR paper was the first to consider the idea of entanglement.
These days, entanglement is a core feature of quantum mechanics, but Einstein had been first to even point out entanglement existed.
This is why the EPR paper was set up, as a thought experiment, because no one had made such an exotic state of matter in 1935.
But in the intervening decades between Einstein's work and Bell's, between 1935 and 1964, entanglement had become a serious topic of study.
By Bell's time, there were reliable ways to make it in the lab.
In fact, Madame Wu had famously reproduced the EPR thought experiment as a real experiment.
But simply doing the EPR experiment in real life isn't enough to tell you which explanation is correct, since both predict the same thing.
But Bell wondered if there was another version of the experiment where the non-local and local theories would have to give a different result.
If you're asking that question and you're playing with the EPR setup, then it's like oh, it's literally not just figuratively a twist, right.
Here's a simplified version of Bell's experiment.
Make your entangled electron and positron again.
But now, instead of just measuring them with a stern Gerlach machine like this, the experimenters get a choice about how to orient their machine.
The three different choices are 0 degrees, 120 degrees, and 240 degrees.
This is the twist on the EPR experiment.
The experimenters get to choose independently.
So the electron might be measured at 0 degrees, while the positron is measured at 240 degrees.
If both experimenters happen to choose the same axis, then we know what happens.
They have to get the opposite result as each other, to conserve spin.
The interesting case is when the experimenters happen to choose different axes.
The number we want to predict here is the disagreement rate, the probability that the electron's result is different from the positron's.
First, let's see what quantum mechanics predicts for this number.
Let's say the electron is measured with the zero degree axis and comes out as plus.
Now that its spin has collapsed to positive Z, the positron spin needs to instantly collapse to be negative Z.
This is the non-local part of quantum mechanics.
But what happens when this positron is measured by a machine tilted at 120 degrees?
You can see that the positron spin is already almost facing the plus end of the machine, so it's much more likely to go to plus.
In fact, there's a 75% chance it goes to plus and only a 25% chance it goes to minus.
So the disagreement rate is 25%.
And we can show that for any two different axes the experimenters select, the geometry is analogous.
So they all have this same disagreement rate of 25%.
Anytime the experimenters choose different axes, they will get the same outcome 75 of the time and different outcomes 25 of the time.
Now let's consider the local hidden variable alternative theory.
The particles here are on a mission.
Their aim to make you believe that they're acting according to quantum mechanics, when really they're acting locally.
Now we are anthropomorphizing, but I think it is really useful to just imagine them this way.
We're trying to figure out if it's always possible for them to use a hidden variable theory to get the same experimental outcomes as Copenhagen quantum mechanics.
Or if in this new situation, our scheming particles won't be able to fool us.
You can think of it like this.
Each of the particles is going to be asked one of three possible questions, and they need to decide on their answer while they're still together so that they can coordinate on their strategy.
When they're done figuring out a plan for how they would answer any of the three questions, they pack away those hidden variables into three sealed envelopes for each particle.
The question is what strategies should our sneaky particles take to make people believe that they're following quantum mechanics?
Remember, quantum mechanics predicted a disagreement rate of 25, and so our particles want to match that.
Whenever they happen to be asked different questions, their answer needs to disagree about 25 of the time.
So, what's the best strategy?
Well, there's actually only two things that they really can do.
The first strategy is this.
The electron answers the same way for each of its axes and the positron answers in the opposite way.
Let's say the electron answers with minus and the positron with plus.
But this is a terrible idea because, whatever two different axes the experimenters happen to choose, the disagreement rate is 100, which is very different from 25.
And so that strategy doesn't work.
But there's only one other strategy that the particles could use.
Instead of the electron doing exactly the same thing for all three axes, it does the same thing for any two of its axes, and then something different for the last one.
Let's just say, for example, that it does this and then the positron does the opposite.
This is just one example, but it turns out for all possible strategies like this, the disagreement rate is going to be the same.
Let's imagine first that the experimenter who's measuring the electron happens to measure it in the 120 degree axis and it gets the answer minus.
They make this choice a third of the time.
And now to calculate the disagreement rate, we need to see what happens when the experimenter who's measuring the positron happens to measure a different axis from this one.
So one of these two.
But in either one of these cases, the positron is also minus.
And so the two answers agree with each other.
And so the answers have no disagreement.
And so we can multiply this by zero.
But two thirds of the time the experimenter who's measuring the electron will happen to measure it in one of the other two axes.
Let's say this one.
The experimenter measuring the polytron will measure in one of these two axes, but you can see that they only pick an axis.
That disagrees half of the time.
That's one third, which is roughly equal to 33%, which is a different number from the quantum one.
When our scheming local particles are interrogated, their answers for different questions match just a little too often.
They simply can't fake the results of Copenhagen quantum mechanics.
So Bell's proof showed that non-local and local theories make different predictions about how often the two results will disagree when the experimenters measure different axes.
Non-local quantum mechanics predicts disagreement only 25% of the time.
Local hidden variables predicts disagreement at least 33% of the time.
So to find out if there really is a local hidden variable theory, you just need to do the experiment.
Okay so welcome at the Institut d'Optique.
You are here in the place where Alain Aspect performed, 40 years ago, his experiments on the measurement of Bell's inequalities, and here are some of the original pictures of this experiment, which was much more challenging than it is today, and it was a real experimental to the force.
So this is one of the original This is one of the polarizers.
It looks like that now.
So only on this small breadboard, right?
It's our main source and this beam is directed towards this element, which is the key element of the setup.
So it's a pair of crystals that produces pairs of entangled photons.
We will produce a pair of entangled photons and both are propagating along each of these two arms.
They are separated.
So here we have the two detection arms and we can rotate the half-wave plate to change the orientation of the measurement basis.
The experiment that we did with light was a little bit different from the one we described earlier with electrons and protons.
So I'm going to explain how they correspond.
Here's a little diagram of the photon experiment.
So first we have this element that makes the entangled pair of particles.
That's these two.
Then the entangled particles go off on separate arms of the experiment.
In our previous experiment we could decide the direction that we're going to measure the particles in by rotating the Stern-Gerlach machines.
And in this experiment, it's actually really similar.
So we have these two polarizers that we're able to rotate independently, and that is going to decide which direction these particles are measured in.
And so this experiment with light is completely equivalent to Bell's experiment.
Bell expected that the experiments would show that the predictions of quantum mechanics were correct and that, you know, there was some kind of non-locality in nature.
Before the first Bell test was done, John Bell said in view of the general success of quantum mechanics, it's very hard for me to doubt the outcome of such experiments.
You didn't expect quantum physics to be wrong because who would bet against quantum physics?
You'd have to be crazy.
Remember, the two different outcomes for Bell's theorem depend on how often two different measurement axes are going to have results that disagree with each other.
Here's how we measure that disagreement rate.
First, we're gonna start with both of the measurements being in the same direction.
And now we expect that these two are always gonna disagree with each other because they have opposite spins.
So we're gonna create a bunch of entangled particles and find out how many of them disagree with each other per second.
This is going to give us a measure of the total number of particles coming per second.
That's because this device is making loads and loads of entangled particles, and so we just need to know how many of them are coming at a time.
Then we rotate one of the axes and now we measure the number of disagreeing pairs per second.
And then dividing these two will give us the disagreement rate.
And remember, quantum mechanics predicts that the disagreement rate will only be a quarter, whereas local hidden variables expects this number to be a third.
So I started at 2000, right?
And now I have 500.
So that really works pretty well.
We did do this experiment again and the number we got very much agreed with quantum mechanics.
But this is one of the most misunderstood experiments in all of physics.
You will find in all sorts of physics textbooks and papers and whatnot, that what Bell's theorem proves is that it rules out local hidden variables or local realism.
John Bell said that was an error.
He said it's really quite remarkable how many people make that error.
I always get confused at the conclusion of Bell's theorem because there's a lot of people who say like okay it rules out hidden variables or things have to be non-local or whatever but what do you think?
Yeah, I think it is super confusing, and when I first learned Bell's theorem, I was told that it rules out local hidden variables.
I've heard this other argument that it sort of disproves either locality or realism.
If you say, okay, it means that you give up local realism.
And so that means you somehow have a choice between giving up locality and giving up realism.
If you're giving up realism, realism about what?
Like, you gotta tell me.
Because, like, for most definitions of that word, you'd also be giving up locality.
So what the hell are you saving?
Like, I just don't...
Yeah, it's a really deep misunderstanding that shows up in almost every single textbook on the subject.
So what does Bell's theorem really prove?
Well, here's the logic.
Start by assuming locality for the entangled particles.
Using the EPR argument.
The only way for them to coordinate their outcomes is using local hidden variables.
Then Bell's proof showed that local hidden variables predict an incorrect experimental result.
Therefore, the assumption of locality must have been wrong.
We're obliged to invoke something like actions going faster than light from one place to another.
The EPR paper by itself had shown that the Copenhagen interpretation is non-local, which is why Einstein thought there must be an alternative way to describe the experiment that is local.
But Bell's theorem says that's not true.
Any theory that correctly describes this experiment must be non-local.
But I still, I would hesitate to say that that means that Einstein was wrong, right?
Because what I would say is this shows that Einstein was right to be concerned about all of this.
People often claim that Einstein's problem was that he simply couldn't accept quantum mechanics, but it was only because he refused to shut up and calculate that he discovered two of the most important aspects of quantum mechanics entanglement and non-locality.
The heart of the debate between Einstein and Bohr was about whether there was a problem, whether there was something to be concerned about.
And the major concern that Einstein brought to the table from the beginning was about locality.
But you know, what Bell showed was, oh yeah, all that stuff that Einstein was concerned about about locality.
He was completely right to be worried about it.
We have a problem.
If these particles really are acting nonlocally, this should cause paradoxes, shouldn't it?
Well, it does, but the paradox seems to be surprisingly tame.
Imagine you and your friend are measuring a pair of entangled particles.
Suppose an observer sees you measure yours first and then your friend measures hers.
That observer thinks that you collapsed the overall state of both particles, and your friend just finds out the result when she measures.
But another observer will see the situation in reverse.
They see her measure first and then you.
To them, it was her measurement that caused the collapse, not yours.
But who's right?
Which measurement was the cause of the collapse, and which was the effect?
It seems to depend on your frame of reference.
This paradox is worrying, but it isn't as bad as the usual faster than light paradoxes.
In relativity.
If you can communicate faster than light, then you can exploit how different observers disagree about timing.
If your friend who's on a rocket sends you an instant message and you send an instant message back in some frames of reference, your message can arrive before she even sent the first one.
If your message says don't send your original message, and so she doesn't, then you've got yourself a paradox.
What prompted you to send this message if she never sent you anything in the first place?
Quantum mechanics sidesteps these paradoxes through a fundamental constraint.
The outcomes are random, so you can't send messages faster than light.
When you measure your particle, you get a plus or a minus completely at random.
Your friend measuring their particle also gets a random result.
Now the results will be correlated, but they're still completely random.
So there's no way to send any faster than light message in this way.
That's what prevents us from sending messages back in time using quantum mechanics.
So quantum mechanics is non-local, but it doesn't lead to the sort of catastrophic paradoxes you might expect from relativity.
But it's an uneasy truce.
Quantum mechanics may not break the letter of relativity's laws, but it certainly violates the spirit.
And nonlocality isn't the only troubling thing about quantum mechanics.
The Copenhagen interpretation still doesn't explain what an electron is really doing and why it acts so differently when measured.
Despite this, many physicists took Bell's theorem to mean that the Copenhagen interpretation was right all along.
Bell himself rejected this.
He spent the rest of his life championing alternative interpretations of quantum mechanics, including the hidden variable interpretation called pilot wave theory or Bohmian mechanics.
Bell's theorem doesn't rule this interpretation out, because the pilot wave theory is non-local, just like the Copenhagen interpretation.
It was Bell's theorem and Bell's subsequent tireless work that made studying the meaning of quantum mechanics respectable again.
He showed that mere armchair philosophy and thought experiments can have real consequences in physics.
We need to be teaching quantum physics in a different way.
We need to be teaching Bell's theorem in a different way.
We do often teach Bell's theorem to physics students and it's taught us something that rules out local hidden variables.
That's just not true.
Bell's theorem, you know, says that quantum physics is in very serious tension with relativity.
On the issue of locality,
John Bell passed away suddenly at the age of 62.
He didn't know it, but he had been nominated for the Nobel Prize just a year earlier.
In a talk he gave in Geneva in January 1990, he said, I think you're stuck with the non-locality.
I don't know any conception of locality which works with quantum mechanics.
That was eight months before he died.
So pretty much his last word on the subject.
And so that's it.
There really are faster than light influences in the universe.
Bell's theorem proves it.
But maybe there is a way out.
There is another way to interpret quantum mechanics that's even more bizarre than the Copenhagen interpretation.
Imagine the EPR thought experiment again.
We can think of the entangled state as being in a superposition of the electron being up and the positron down, and the electron being down and the positron being up.
In the Copenhagen interpretation.
When you measure a particle and you get only one result, say plus, the other part of the superposition collapses.
But in our examples, we've seen that measurement collapse seems to be the source of non-locality.
So why don't we just get rid of collapse altogether?
This is what the Many Worlds Interpretation of Quantum Mechanics proposes.
When you measure a particle, instead of you collapsing the particle to one outcome, both outcomes happen and there's two parallel versions of you.
Who sees each outcome?
You have become entangled with your particle because your state depends on what the particle is doing.
It sounds strange, but there's one huge benefit of this interpretation.
When your friend is about to measure her electron, your positron doesn't need to rush to tell the electron what the answer will be.
There are already two versions of the electron and they contain the right answer for each version of her.
There was no need for faster-than-light communication to explain the EPR experiment.
But how is that possible?
Doesn't Bell's theorem prove that the two particles must communicate faster than light?
Well, in Bell's proof, we assumed that all measurements have just one outcome.
But that assumption just isn't true in many worlds.
This means that technically, that proof doesn't even apply in the many worlds case.
So, is many worlds local?
In one sense no because, just like in Copenhagen, quantum mechanics entangled pairs can be separated by a huge distance and still share their state.
However, it is local, unlike Copenhagen, in the sense that these faraway entangled particles do not influence each other faster than light.
Many worlds obeys Einstein's universal speed limit.
But is it really worth accepting that there are many versions of you in parallel universes just to recover locality?
Well, locality isn't the only reason Many Worlds has become more and more popular.
I also really like many worlds.
Because Copenhagen never sits right.
And when you start telling the story of what happens at measurement, it's like well, what is a measurement?
It's when you have this quantum system and there's some other system which is much larger.
But it always feels a little bit arbitrary.
Whereas this argument that every time two quantum particles are interacting, their wave functions are essentially combining and becoming entangled.
That to me feels more consistent.
Yes, I think that's right.
What do you think are like the problems with many worlds?
The biggest problem is, I think, people's struggle to deal with sort of the infinity that that brings forth.
For sure.
But I don't know that that's necessarily an argument against it.
Just because it's hard to imagine doesn't mean it's not what's happening.
If many worlds is right, everything changes.
The conflict between quantum mechanics and relativity vanishes.
Physicists have been struggling for decades to unite quantum mechanics with general relativity to build a theory of quantum gravity.
And maybe we've been failing because we've been trying to marry relativity to a non-local theory.
But if quantum mechanics ultimately turns out to be local, well then Einstein's dream of a local description of reality might not be dead after all.
