## Friday, March 24, 2017

### Learning about the laws of physics isn't a "yes we can" pissing contest

After Sabine Hossenfelder wrote her critique of "the world is a simulation" paradigm, I was a bit jealous about one apparent phenomenon: that her readers seemed to agree with her. Well, it didn't last long. After Scott Aaronson vented his absolutely stupid ideas about the same problem, many of his computer-science-worshiping but otherwise uneducated readers were apparently redirected to Hossenfelder's blog and started to give her a hard time.

The most obnoxious troll that repeatedly posted at Backreaction is nicknamed _Shorty, a man from the British Columbia who loves his air gun, guitar, and video games. For some reasons, this self-evident mediocre know-nothing thinks that it's very important for the world to hear what he thinks about the character of the physical law. It wouldn't be too hard to predict what an interaction between a physicist, even one such as Hossenfelder, and a stupid yet aggressive man who is "into the computer games" is going to look like.

Our Shorty moron wrote:
You're basing your opinion on what you know, and/or what you think you know.

[Computers are getting more powerful.]

[Computers can simulate everything.] There is absolutely nothing about the nature of quantum mechanics that would be restricted by a binary computer. [Bell's theorem isn't an obstacle, either.]

No need to look up anything. You think because you can't do something, nobody can. Not even someone with more knowledge and technology than you.

As I said, you just don't get it. Your argument is no different than someone from 15,000 years ago thinking it is impossible for people to fly, so nobody will ever fly. And yet, planes. Da plane, boss!
With the help of some brackets and a careful selection, I've shortened Shorty's tirades by more than one order of magnitude – but I hope that this operation meant to save your time hasn't robbed you of any contents you should know. It's just incredible how much redundant garbage the stupid people have to write to express a trivial and well-known fact e.g. that "computers are getting better".

You know, Shorty, none of your shouting can make the hypothesis that the Universe is a simulation any more likely. As Hossenfelder repeatedly had to point out, Shorty was just talking about a subject – namely physics – that he has absolutely no clue about. It's very clear that even Sabine Hossenfelder becomes an Albert Einstein – if not an Albrecht Dreistein or Donald Rammstein – in comparison.

OK, let's look at the points as I have summarized them.
Shorty to Sabine: You're basing your opinion on what you know, and/or what you think you know.
I had to laugh because I've heard this criticism many times as well and I responded basically just like Hossenfelder did:
So what do you recommend? That I base my opinion on what I don't know?
It's even plausible that Hossenfelder has learned this answer from me – but I am not insisting on taking the credit. ;-) You know, the strategy of Shorty's criticism is that people aren't omniscient. And contemporary people – and physicists – probably know less than the future people – and physicists – will know. So you can use this imperfection to dismiss any opinion that a person – especially a physicist – can tell you today.

Nice.

A point that Shorty overlooks is that he is a person as well – and he is also a contemporary person. He is not a physicist but it makes things worse, not better, as we can easily see. He is arguing with Sabine Hossenfelder and both of them are people who are currently alive. So none of them can legitimately act as if he were a man from the future who already knows more things – or who knows a more accurate truth about the world – just because of a comparison of dates. The dates are the same – otherwise they couldn't have interacted with one another – which means that any party's claim of superiority based on the dates is obviously invalid.

If we want to learn something about the world, we must build on things that we do know, not on things that we don't know (e.g. we don't know yet). This makes our research harder and/or less reliable but it's the best thing we can do. If we start to base our thoughts about the world on things that we don't know at all, e.g. on pure fantasies, speculations, and a wishful thinking, we are pretty much guaranteed to end up with much worse results!

Winston Churchill has said
Democracy is the worst form of government, except for all the others.
In the same way, we may say:
Applying the scientific method on the insights that we know now is the worst method to learn about Nature, except for all the other methods.
OK, I hope that if you think about it, you will agree that a person whose criticism is based on the assertion that his antagonist is "building on things he or she knows" is an imbecile.
[Computers are getting more powerful.]
He spends a lot of time by saying that Hossenfelder's Dell is better than computers 50 years ago and in 5 years, her computer will be OK for a trash bin, too. Does he really believe that she doesn't know that computer technology is making progress? If he does, it's probably because the insight that computers are making progress is one of the most intellectually difficult theories that his peabrain has been capable of understanding.

Of course Hossenfelder knows that. Every sane enough person knows about the progress of computers.

But this progress has absolutely nothing to do with the question what are the laws underlying our Universe and the history of all the objects that were needed for the things we know and observe to emerge. The likely progress of computers affects answers to questions about the future while physics and cosmology is basically the accumulation of all timeless questions and questions about the past. These two sets are not the same, especially because${\rm Future} \neq {\rm Past}.$ Shorty clearly suffers from a shortage of brainpower that is needed to understand this inequality.

Even if the computers and programs in the future were able to create a highly realistic caricature of our Universe – those $$10^{60}$$ Planck lengths in radius, with the Planckian resolution etc. (and we can't really prove that the answer to this question is "Yes") – it will simply not change the truth about the history of our actual Universe. Can you understand this simple point – that understanding the pre-existing world is something completely different than creating a new world? Hossenfelder isn't boasting that she can build – or prevent the construction of – some very powerful computers and programs. She is saying something about the world that already existed before she was born.

Folks like Craig Venter have already started to produce organisms that carry an artificially engineered DNA code (even though, so far, Venter et al. have been plagiarizing Nature way too much). Sometime in the future, this discipline may run through a very successful week in which dozens of very complex organisms and their DNA will be created from scratch. But when it happens (even if it happens), this achievement won't mean that Craig Venter was created by the Lord in seven days as described in the Bible. Can't you see that these questions are completely different?

The possible fact that something will become easy in the future doesn't mean that it was likely to happen in the past.

I guess that the anthropic people may be tempted to make this trivial mistake but I still believe that most people in science – and even most people who have any common sense left – agree with me that it is an absolutely elementary, childish fallacy to confuse "what can be done [by humans or other intelligent entities] in the future" with "what has happened in the past". These just can't be equivalent questions due to the very evolution, due to the arrow of time, due to the purpose of physics to give a simpler explanation for the more complex observations.

The purpose of physics is to find simple and rigid enough theories that can explain the data. It means that we want to assume as little as possible – yet be able to predict (and only correctly predict) as much as possible. So physics (plus cosmology and other "historical sciences") are really describing how things arise from nothing or almost nothing. If you start with a world which has far more advanced computers (or far more intelligent beings) than what we have today, you are not making any progress in physics at all – because you have basically "inserted" more than what you "extracted" from your theory. So your contribution to science is negative. Your theory has created a one-time profit that is smaller than the original investment so you are a loser.

If I have to give you a very specific example, atomic and molecular physics (quantum mechanics) calculates millions of spectral lines of atoms and molecules. All these data are correctly explained by – reduced to – a few assumptions about the nuclei, electrons, and their interactions. What we observe follows from the assumptions. But if you say that our world is really a simulation, there is no need for the spectral lines to agree with those calculated from quantum mechanics because you just said that quantum mechanics with the simple interactions doesn't fundamentally hold. Instead, it's only being emulated. So you must probably say that your computer tries to simulate quantum mechanics so that all the previous successes of quantum mechanics are reproduced by your theory. But you haven't really explained why. A simulation may use completely different and independent spectral lines for all atoms and molecules. It could make a computer game cooler, simpler, or more general, or better in some other way. Like filmmakers in Hollywood, computer game programmers aren't usually obsessed with the 100% accurate validity of the laws of physics.

So while the conventional physical theory which assumes that quantum electrodynamics is "really" right does explain why the spectral lines are calculable by those equations, your simulation hypothesis simply doesn't explain why. The usual theory does reduce these patterns to a simple assumption – the equations really hold – while your theory doesn't – the assumption about a computer simulation implies that some of the patterns are almost certainly violated. So your hypothesis is not on part with physics as we know it. Instead, it's on par with any other superstition, religion, fantasy, or fairy-tale that someone invented and that is also unlikely to produce the predictions that are needed. If you pretend that your theory is better than what a 3-year-old kid babbled about the dwarfs controlling the Earth last night, it just means that you are abusing children. Your theory is not better.

Next, we have the theme of the anti-quantum zealots again.
[Computers can simulate everything.] There is absolutely nothing about the nature of quantum mechanics that would be restricted by a binary computer. [Bell's theorem isn't an obstacle, either.]
The Shorty troll clearly wants to defend a classical simulation of our quantum world. As I previously wrote, this effort places him to the stupidest layer of the already stupid community of people who defend the idea that the Universe may be a simulation.

No, Shorty, it isn't possible to design a computer of the ordinary type – which we call the classical computer – whose behavior would be equivalent to a quantum system. We don't need to know whether the CPU power of computers will jump 10-fold or a quadrillion-fold. The general defining assumptions of a classical computer are enough: the high CPU power just doesn't help you with the conceptual problems at all. Hossenfelder refers to Bell's theorem and says that if the computer were non-local, perhaps it could work.

It couldn't work with a non-local classical computer, either. Bell's theorem is just one childish exercise showing that quantum mechanics is different from any previously conceivable theory in classical physics. It shows that local classical theories just can't predict the same statistical results for two spins as quantum mechanics – which also agrees with the observations. But there are infinitely many similar theorems that exclude classes of non-local classical theories as well, pick e.g. the 2007 work by Zeilinger et al. for a very specific example.

Quantum mechanics works fundamentally differently than classical physics – and therefore classical computers. Its predictions are observer-dependent – they are calculations of probabilities of future outcomes of observations calculated from something that an observer subjectively (but correctly) considers his observations in the past. This observer-dependent, complementary character of reality just cannot be captured by any classical theory i.e. any classical computer, whether its laws are local or not. Physics has simply learned that the laws of the Universe are local and non-realist. Just learn what these laws say and try to understand why there can't be fundamentally different laws that would nevertheless be equally successful.

I don't want to expand this portion too much because Shorty obviously doesn't have a clue about quantum mechanics so a lecture on quantum mechanics would be a cannon against ant, and I have written some 700 too many lectures on foundations of quantum mechanics, anyway. Quantum mechanics is so different that peabrains of Shorty's caliber have no chance to ever understand it – or even to understand that there could be something going so dramatically beyond their abilities that they have no chance to ever understand it. A discussion about quantum mechanics with Shorty is approximately as meaningless as a discussion about quantum mechanics with an angry dog.

That's why I want to return to more general points. Shorty also wrote:
No need to look up anything. You think because you can't do something, nobody can. Not even someone with more knowledge and technology than you.
If Hossenfelder were using this way of thinking, it would indeed be fallacious. Some people think that nobody can do something because they can't do it. However, in this case, it's exactly the other way around: Hossenfelder has used a completely different argument (namely the scientific method: she extracted some general predictions from the "simulation hypothesis" and falsified them i.e. showed that they disagree with the observations) while it's Shorty himself who has used the fallacious argument, who thinks that because he can't do something, nobody can. What is this something? It's a proof that the Universe isn't a simulation. So the relevant answer to a Shorty troll would be:
Dear Shorty troll, you think that because you can't prove that the "Universe is a simulation" hypothesis is incorrect, nobody else can prove it, either. But this implication is logically invalid.

Your accusations that your antagonist resorts to a fallacy – even though you are the only one who actually does so – is known as the psychological projection.
The troll cannot prove that the simulation hypothesis is false. But we can prove it, at least "statistically" with a high probability extremely close to 100%.

Physics is a science and science generally wants to find correct answers to questions. A funny thing about Yes/No questions in science is that both Yes and No are possible. A scientist, like an idealized clever and honest human being, must use all existing or newly invented intelligent methods to accumulate all accessible evidence in favor of Yes and all accessible evidence in favor of No and impartially compare them.

When it is done with the "simulation hypothesis", the conclusion is that it is a hopeless hypothesis with no realistic chance to be right. The scientific evidence just strongly points to the answer "No". Impressively realistic computer games of the future can't change anything about this conclusion – at most, they are answers to completely different questions.

To fight against the evidence in favor of "No", a fanatical proponent may declare that all the evidence is irrelevant and those who have understood that and why the answer is "No" are analogous to all sorts of unpopular people or people who were wrong in the past. Or he can say "Yes, a simulation can" (where the world "can" inflates an arbitrarily tiny probability into "effectively 100%") while ignoring that "No, the simulation probably won't". But that's simply not a method to rationally, let alone scientifically, investigate or argue. Sciences compares possible answers and says that one of them seems more likely than others. It doesn't try to mindlessly rationalize a privileged answer that was picked as the winner to start with.

Specifically, Shorty wrote:
As I said, you just don't get it. Your argument is no different than someone from 15,000 years ago thinking it is impossible for people to fly, so nobody will ever fly. And yet, planes. Da plane, boss!
When Shorty uses this widespread talking point, he should at least learn it correctly. The statement that machines heavier than air would never fly is indeed famous. But it wasn't made 15,000 years ago. Instead, it was made in 1895 by Lord Kelvin, one of the most admired physicists (and, sociologically, probably the world's #1 physicist) of the late 19th century. And this statement – which wasn't really backed by any true physics argumentation or calculation – was proven wrong just 8 years later. At least, with his silly no-go theorem for airplanes, Lord Kelvin showed that he wasn't "not even wrong". ;-)

I would personally say that his assertion had very little to do with physics. After all, even folks who lived centuries ago such as Leonardo Da Vinci basically knew how heavy flying machines could be constructed – his machines could have been more inspired by birds than the future helicopters but the birds are ultimately counterexamples to Kelvin's assertion, too, which is what makes it so puzzling why he would ever say such a thing. Kelvin was just bullšiting, voicing his personal prejudices.

Kelvin was wrong in several other cases. But some of the others showed his good physics background. For example, Kelvin has disagreed with Darwin concerning the age of the Earth. Darwin was basically right (the age was above 300 million years because of erosion of chalk) while Kelvin calculated a much shorter age, 100 million years, basically from the energy that the Sun could have stored (while misunderstanding the very concentrated nuclear energy in the Sun). Kelvin was wrong in this one as well but I think that good physicists would still agree that except for the neglected form of energy, his argument was of the type that a physicist shouldn't be ashamed of.

Anyway, let's return to "da planes". Kelvin was wrong about the impossibility of airplanes. But there's no analogous proof that Hossenfelder or your humble correspondent are wrong about "the Universe is a simulation". So these two things simply aren't equivalent. Just comparing a disputed proposition $$P$$ to a completely different proposition $$W$$ that is known to be wrong doesn't prove that $$P$$ is wrong, does it? Such a comparison is nothing else than demagogy – and if Shorty's antagonist were a demagogue as well, she or he could offer a fully analogous but opposite metaphor.

To summarize, I would say that individuals like Shorty are morons and bullies who have no clue about the questions they are trying to influence but who are eager to terrorize someone who gives answers – or evidence for answers – that they dislike. They think that the arguments may be replaced with the linking of their antagonists to some unpopular or wrong people in the past, with the declaration that the antagonists are heretics. It's heretical to say that the Universe isn't a simulation because such a proposition must surely insult the pride of men who love to play with an air gun and who play video games, right?

Sorry but physics doesn't respect these labels of "heresy" and doesn't consider anything that Shorty has written to be a valid argument that would be relevant for the question whether the Universe may be a simulation. From the viewpoint of physics, Shorty is just a noise-producing bound state of electrons and nuclei – just like a 200-pound pile of feces. In fact, both from the viewpoint of physics and psychology, Shorty is basically indistinguishable from the aforementioned pile. Physics itself only pays attention to the arguments, not to the yelling by Shorties, and the arguments – predictions about discreteness, glitches, and various features that are very likely to exist in any simulated Universe but that demonstrably don't exist in our Universe – speak a rather clear language.

The answer is "No, the Universe almost certainly cannot be a simulation". Get used to it, Shorty, or shoot yourself with your air gun.