I’ve never met someone who believed the Earth was flat. I’ve met a few who believed it was six thousand years old, but not many. Occasionally, I run into crackpots who rail against relativity or quantum mechanics, or more recent discoveries like quarks or the Higgs. But for one conclusion of modern physics, the doubters are common. For this one idea, the average person may not insist that the physicists are wrong, but they’ll usually roll their eyes a little bit, ask the occasional “really?”
That idea is dark matter.
For the average person, dark matter doesn’t sound like normal, responsible science. It sounds like cheating. Scientists try to explain the universe, using stars and planets and gravity, and eventually they notice the equations don’t work, so they just introduce some new matter nobody can detect. It’s as if a budget didn’t add up, so the accountant just introduced some “dark expenses” to hide the problem.
Part of what’s going on here is that fundamental physics, unlike other fields, doesn’t have to reduce to something else. An accountant has to explain the world in terms of transfers of money, a chemist in terms of atoms and molecules. A physicist has to explain the world in terms of math, with no more restrictions than that. Whatever the “base level” of another field is, physics can, and must, go deeper.
But that doesn’t explain everything. Physics may have to explain things in terms of math, but we shouldn’t just invent new math whenever we feel like it. Surely, we should prefer explanations in terms of things we know to explanations in terms of things we don’t know. The question then becomes, what justifies the preference? And when do we get to break it?
Imagine you’re camping in your backyard. You’ve brought a pack of jumbo marshmallows. You wake up to find a hole torn in the bag, a few marshmallows strewn on a trail into the bushes, the rest gone. You’re tempted to imagine a new species of ant, with enormous jaws capable of ripping open plastic and hauling the marshmallows away. Then you remember your brother likes marshmallows, and it’s probably his fault.
Now imagine instead you’re camping in the Amazon rainforest. Suddenly, the ant explanation makes sense. You may not have a particular species of ants in mind, but you know the rainforest is full of new species no-one has yet discovered. And you’re pretty sure your brother couldn’t have flown to your campsite in the middle of the night and stolen your marshmallows.
We do have a preference against introducing new types of “stuff”, like new species of ants or new particles. We have that preference because these new types of stuff are unlikely, based on our current knowledge. We don’t expect new species of ants in our backyards, because we think we have a pretty good idea of what kinds of ants exist, and we think a marshmallow-stealing brother is more likely. That preference gets dropped, however, based on the strength of the evidence. If it’s very unlikely our brother stole the marshmallows, and if we’re somewhere our knowledge of ants is weak, then the marshmallow-stealing ants are more likely.
Dark matter is a massive leap. It’s not a massive leap because we can’t see it, but simply because it involves new particles, particles not in our Standard Model of particle physics. (Or, for the MOND-ish fans, new fields not present in Einstein’s theory of general relativity.) It’s hard to justify physics beyond the Standard Model, and our standards for justifying it are in general very high: we need very precise experiments to conclude that the Standard Model is well and truly broken.
For dark matter, we keep those standards. The evidence for some kind of dark matter, that there is something that can’t be explained by just the Standard Model and Einstein’s gravity, is at this point very strong. Far from a vague force that appears everywhere, we can map dark matter’s location, systematically describe its effect on the motion of galaxies to clusters of galaxies to the early history of the universe. We’ve checked if there’s something we’ve left out, if black holes or unseen planets might cover it, and they can’t. It’s still possible we’ve missed something, just like it’s possible your brother flew to the Amazon to steal your marshmallows, but it’s less likely than the alternatives.
Also, much like ants in the rainforest, we don’t know every type of particle. We know there are things we’re missing: new types of neutrinos, or new particles to explain quantum gravity. These don’t have to have anything to do with dark matter, they might be totally unrelated. But they do show that we should expect, sometimes, to run into particles we don’t already know about. We shouldn’t expect that we already know all the particles.
If physicists did what the cartoons suggest, it really would be cheating. If we proposed dark matter because our equations didn’t match up, and stopped checking, we’d be no better than an accountant adding “dark money” to a budget. But we didn’t do that. When we argue that dark matter exists, it’s because we’ve actually tried to put together the evidence, because we’ve weighed it against the preference to stick with the Standard Model and found the evidence tips the scales. The instinct to call it cheating is a good instinct, one you should cultivate. But here, it’s an instinct physicists have already taken into account.
There’s a saying in physics, attributed to the famous genius John von Neumann: “With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.”
Say you want to model something, like some surprising data from a particle collider. You start with some free parameters: numbers in your model that aren’t decided yet. You then decide those numbers, “fixing” them based on the data you want to model. Your goal is for your model not only to match the data, but to predict something you haven’t yet measured. Then you can go out and check, and see if your model works.
The more free parameters you have in your model, the easier this can go wrong. More free parameters make it easier to fit your data, but that’s because they make it easier to fit any data. Your model ends up not just matching the physics, but matching the mistakes as well: the small errors that crop up in any experiment. A model like that may look like it’s a great fit to the data, but its predictions will almost all be wrong. It wasn’t just fit, it was overfit.
We have statistical tools that tell us when to worry about overfitting, when we should be impressed by a model and when it has too many parameters. We don’t actually use these tools correctly, but they still give us a hint of what we actually want to know, namely, whether our model will make the right predictions. In a sense, these tools form the mathematical basis for Occam’s Razor, the idea that the best explanation is often the simplest one, and Occam’s Razor is a critical part of how we do science.
So, did you know machine learning was just modeling data?
All of the much-hyped recent advances in artificial intelligence, GPT and Stable Diffusion and all those folks, at heart they’re all doing this kind of thing. They start out with a model (with a lot more than five parameters, arranged in complicated layers…), then use data to fix the free parameters. Unlike most of the models physicists use, they can’t perfectly fix these numbers: there are too many of them, so they have to approximate. They then test their model on new data, and hope it still works.
Increasingly, it does, and impressively well, so well that the average person probably doesn’t realize this is what it’s doing. When you ask one of these AIs to make an image for you, what you’re doing is asking what image the model predicts would show up captioned with your text. It’s the same sort of thing as asking an economist what their model predicts the unemployment rate will be when inflation goes up. The machine learning model is just way, way more complicated.
As a physicist, the first time I heard about this, I had von Neumann’s quote in the back of my head. Yes, these machines are dealing with a lot more data, from a much more complicated reality. They literally are trying to fit elephants, even elephants wiggling their trunks. Still, the sheer number of parameters seemed fishy here. And for a little bit things seemed even more fishy, when I learned about double descent.
Suppose you start increasing the number of parameters in your model. Initially, your model gets better and better. Your predictions have less and less error, your error descends. Eventually, though, the error increases again: you have too many parameters so you’re over-fitting, and your model is capturing accidents in your data, not reality.
In machine learning, weirdly, this is often not the end of the story. Sometimes, your prediction error rises, only to fall once more, in a double descent.
For a while, I found this deeply disturbing. The idea that you can fit your data, start overfitting, and then keep overfitting, and somehow end up safe in the end, was terrifying. The way some of the popular accounts described it, like you were just overfitting more and more and that was fine, was baffling, especially when they seemed to predict that you could keep adding parameters, keep fitting tinier and tinier fleas on the elephant’s trunk, and your predictions would never start going wrong. It would be the death of Occam’s Razor as we know it, more complicated explanations beating simpler ones off to infinity.
Luckily, that’s not what happens. And after talking to a bunch of people, I think I finally understand this enough to say something about it here.
The right way to think about double descent is as overfitting prematurely. You do still expect your error to eventually go up: your model won’t be perfect forever, at some point you will really overfit. It might take a long time, though: machine learning people are trying to model very complicated things, like human behavior, with giant piles of data, so very complicated models may often be entirely appropriate. In the meantime, due to a bad choice of model, you can accidentally overfit early. You will eventually overcome this, pushing past with more parameters into a model that works again, but for a little while you might convince yourself, wrongly, that you have nothing more to learn.
So Occam’s Razor still holds, but with a twist. The best model is simple enough, but no simpler. And if you’re not careful enough, you can convince yourself that a too-simple model is as complicated as you can get.
Image from Astral Codex Ten
I was reminded of all this recently by somearticles by Sabine Hossenfelder.
Hossenfelder is a critic of mainstream fundamental physics. The articles were her restating a point she’s made many times before, including in (at least) one of her books. She thinks the people who propose new particles and try to search for them are wasting time, and the experiments motivated by those particles are wasting money. She’s motivated by something like Occam’s Razor, the need to stick to the simplest possible model that fits the evidence. In her view, the simplest models are those in which we don’t detect any more new particles any time soon, so those are the models she thinks we should stick with.
I tend to disagree with Hossenfelder. Here, I was oddly conflicted. In some of her examples, it seemed like she had a legitimate point. Others seemed like she missed the mark entirely.
Talk to most astrophysicists, and they’ll tell you dark matter is settled science. Indeed, there is a huge amount of evidence that something exists out there in the universe that we can’t see. It distorts the way galaxies rotate, lenses light with its gravity, and wiggled the early universe in pretty much the way you’d expect matter to.
What isn’t settled is whether that “something” interacts with anything else. It has to interact with gravity, of course, but everything else is in some sense “optional”. Astroparticle physicists use satellites to search for clues that dark matter has some other interactions: perhaps it is unstable, sometimes releasing tiny signals of light. If it did, it might solve other problems as well.
Hossenfelder thinks this is bunk (in part because she thinks those other problems are bunk). I kind of do too, though perhaps for a more general reason: I don’t think nature owes us an easy explanation. Dark matter isn’t obligated to solve any of our other problems, it just has to be dark matter. That seems in some sense like the simplest explanation, the one demanded by Occam’s Razor.
At the same time, I disagree with her substantially more on collider physics. At the Large Hadron Collider so far, all of the data is reasonably compatible with the Standard Model, our roughly half-century old theory of particle physics. Collider physicists search that data for subtle deviations, one of which might point to a general discrepancy, a hint of something beyond the Standard Model.
While my intuitions say that the simplest dark matter is completely dark, they don’t say that the simplest particle physics is the Standard Model. Back when the Standard Model was proposed, people might have said it was exceptionally simple because it had a property called “renormalizability”, but these days we view that as less important. Physicists like Ken Wilson and Steven Weinberg taught us to view theories as a kind of series of corrections, like a Taylor series in calculus. Each correction encodes new, rarer ways that particles can interact. A renormalizable theory is just the first term in this series. The higher terms might be zero, but they might not. We even know that some terms cannot be zero, because gravity is not renormalizable.
The two cases on the surface don’t seem that different. Dark matter might have zero interactions besides gravity, but it might have other interactions. The Standard Model might have zero corrections, but it might have nonzero corrections. But for some reason, my intuition treats the two differently: I would find it completely reasonable for dark matter to have no extra interactions, but very strange for the Standard Model to have no corrections.
I think part of where my intuition comes from here is my experience with other theories.
One example is a toy model called sine-Gordon theory. In sine-Gordon theory, this Taylor series of corrections is a very familiar Taylor series: the sine function! If you go correction by correction, you’ll see new interactions and more new interactions. But if you actually add them all up, something surprising happens. Sine-Gordon turns out to be a special theory, one with “no particle production”: unlike in normal particle physics, in sine-Gordon particles can neither be created nor destroyed. You would never know this if you did not add up all of the corrections.
String theory itself is another example. In string theory, elementary particles are replaced by strings, but you can think of that stringy behavior as a series of corrections on top of ordinary particles. Once again, you can try adding these things up correction by correction, but once again the “magic” doesn’t happen until the end. Only in the full series does string theory “do its thing”, and fix some of the big problems of quantum gravity.
If the real world really is a theory like this, then I think we have to worry about something like double descent.
Remember, double descent happens when our models can prematurely get worse before getting better. This can happen if the real thing we’re trying to model is very different from the model we’re using, like the example in this explainer that tries to use straight lines to match a curve. If we think a model is simpler because it puts fewer corrections on top of the Standard Model, then we may end up rejecting a reality with infinite corrections, a Taylor series that happens to add up to something quite nice. Occam’s Razor stops helping us if we can’t tell which models are really the simple ones.
The problem here is that every notion of “simple” we can appeal to here is aesthetic, a choice based on what makes the math look nicer. Other sciences don’t have this problem. When a biologist or a chemist wants to look for the simplest model, they look for a model with fewer organisms, fewer reactions…in the end, fewer atoms and molecules, fewer of the building-blocks given to those fields by physics. Fundamental physics can’t do this: we build our theories up from mathematics, and mathematics only demands that we be consistent. We can call theories simpler because we can write them in a simple way (but we could write them in a different way too). Or we can call them simpler because they look more like toy models we’ve worked with before (but those toy models are just a tiny sample of all the theories that are possible). We don’t have a standard of simplicity that is actually reliable.
From the Wikipedia page for dark matter halos
There is one other way out of this pickle. A theory that is easier to write down is under no obligation to be true. But it is more likely to be useful. Even if the real world is ultimately described by some giant pile of mathematical parameters, if a simple theory is good enough for the engineers then it’s a better theory to aim for: a useful theory that makes peoples’ lives better.
I kind of get the feeling Hossenfelder would make this objection. I’ve seen her argue on twitter that scientists should always be able to say what their research is good for, and her Guardian article has this suggestive sentence: “However, we do not know that dark matter is indeed made of particles; and even if it is, to explain astrophysical observations one does not need to know details of the particles’ behaviour.”
Ok yes, to explain astrophysical observations one doesn’t need to know the details of dark matter particles’ behavior. But taking a step back, one doesn’t actually need to explain astrophysical observations at all.
Astrophysics and particle physics are not engineering problems. Nobody out there is trying to steer a spacecraft all the way across a galaxy, navigating the distribution of dark matter, or creating new universes and trying to make sure they go just right. Even if we might do these things some day, it will be so far in the future that our attempts to understand them won’t just be quaint: they will likely be actively damaging, confusing old research in dead languages that the field will be better off ignoring to start from scratch.
Because of that, usefulness is also not a meaningful guide. It cannot tell you which theories are more simple, which to favor with Occam’s Razor.
Hossenfelder’s highest-profile recent work falls afoul of one or the other of her principles. Her work on the foundations of quantum mechanics could genuinely be useful, but there’s no reason aside from claims of philosophical beauty to expect it to be true. Her work on modeling dark matter is at least directly motivated by data, but is guaranteed to not be useful.
I’m not pointing this out to call Hossenfelder a hypocrite, as some sort of ad hominem or tu quoque. I’m pointing this out because I don’t think it’s possible to do fundamental physics today without falling afoul of these principles. If you want to hold out hope that your work is useful, you don’t have a great reason besides a love of pretty math: otherwise, anything useful would have been discovered long ago. If you just try to model existing data as best you can, then you’re making a model for events far away or locked in high-energy particle colliders, a model no-one else besides other physicists will ever use.
I don’t know the way through this. I think if you need to take Occam’s Razor seriously, to build on the same foundations that work in every other scientific field…then you should stop doing fundamental physics. You won’t be able to make it work. If you still need to do it, if you can’t give up the sub-field, then you should justify it on building capabilities, on the kind of “practice” Hossenfelder also dismisses in her Guardian piece.
We don’t have a solid foundation, a reliable notion of what is simple and what isn’t. We have guesses and personal opinions. And until some experiment uncovers some blinding flash of new useful meaningful magic…I don’t think we can do any better than that.
During the pandemic, some conferences went online. Others went dormant.
Everysummer before the pandemic, the Niels Bohr International Academy hosted a conference called Current Themes in High Energy Physics and Cosmology. Current Themes is a small, cozy conference, a gathering of close friends some of whom happen to have Nobel prizes. Holding it online would be almost missing the point.
A particularly special Current Themes means some unusually special guests. Our guests are usually pretty special already (Gerard t’Hooft and David Gross are regulars, to just name the Nobelists), but this year we also had Alexander Polyakov. Polyakov’s talk had a magical air to it. In a quiet voice, broken by an impish grin when he surprised us with a joke, Polyakov began to lay out five unsolved problems he considered interesting. In the end, he only had time to present one, related to turbulence: when Gross asked him to name the remaining four, the second included a term most of us didn’t recognize (striction, known in a magnetic context and which he wanted to explore gravitationally), so the discussion hung while he defined that and we never did learn what the other three problems were.
At the big 100th anniversary celebration earlier in the spring, the Institute awarded a few years worth of its Niels Bohr Institute Medal of Honor. One of the recipients, Paul Steinhardt, couldn’t make it then, so he got his medal now. After the obligatory publicity photos were taken, Steinhardt entertained us all with a colloquium about his work on quasicrystals, including the many adventures involved in finding the first example “in the wild”. I can’t do the story justice in a short blog post, but if you won’t have the opportunity to watch him speak about it then I hear his book is good.
An anniversary conference should have some historical elements as well. For this one, these were ably provided by David Broadhurst, who gave an after-dinner speech cataloguing things he liked about Bohr. Some was based on public information, but the real draw were the anecdotes: his own reminiscences, and those of people he knew who knew Bohr well.
The other talks covered interesting ground: from deep approaches to quantum field theory, to new tools to understand black holes, to the implications of causality itself. One out of the ordinary talk was by Sabrina Pasterski, who advocated a new model of physics funding. I liked some elements (endowed organizations to further a subfield) and am more skeptical of others (mostly involving NFTs). Regardless it, and the rest of the conference more broadly, spurred a lot of good debate.
Today, we’d call Leibniz a mathematician, a physicist, and a philosopher. As a mathematician, Leibniz turned calculus into something his contemporaries could actually use. As a physicist, he championed a doomed theory of gravity. In philosophy, he seems to be most remembered for extremely cheaty arguments.
Free will and determinism? Can’t it just be a coincidence?
I don’t blame him for this. Faced with a tricky philosophical problem, it’s enormously tempting to just blaze through with an answer that makes every subtlety irrelevant. It’s a temptation I’ve succumbed to time and time again. Faced with a genie, I would always wish for more wishes. On my high school debate team, I once forced everyone at a tournament to switch sides with some sneaky definitions. It’s all good fun, but people usually end up pretty annoyed with you afterwards.
People were annoyed with Leibniz too, especially with his solution to the problem of evil. If you believe in a benevolent, all-powerful god, as Leibniz did, why is the world full of suffering and misery? Leibniz’s answer was that even an all-powerful god is constrained by logic, so if the world contains evil, it must be logically impossible to make the world any better: indeed, we live in the best of all possible worlds. Voltaire famously made fun of this argument in Candide, dragging a Leibniz-esque Professor Pangloss through some of the most creative miseries the eighteenth century had to offer. It’s possibly the most famous satire of a philosopher, easily beating out Aristophanes’ The Clouds (which is also great).
Physicists can also get accused of cheaty arguments, and probably the most mocked is the idea of a multiverse. While it hasn’t had its own Candide, the multiverse has been criticized by everyone from bloggers to Nobel prizewinners. Leibniz wanted to explain the existence of evil, physicists want to explain “unnaturalness”: the fact that the kinds of theories we use to explain the world can’t seem to explain the mass of the Higgs boson. To explain it, these physicists suggest that there are really many different universes, separated widely in space or built in to the interpretation of quantum mechanics. Each universe has a different Higgs mass, and ours just happens to be the one we can live in. This kind of argument is called “anthropic” reasoning. Rather than the best of all possible worlds, it says we live in the world best-suited to life like ours.
I called Leibniz’s argument “cheaty”, and you might presume I think the same of the multiverse. But “cheaty” doesn’t mean “wrong”. It all depends what you’re trying to do.
Leibniz’s argument and the multiverse both work by dodging a problem. For Leibniz, the problem of evil becomes pointless: any evil might be necessary to secure a greater good. With a multiverse, naturalness becomes pointless: with many different laws of physics in different places, the existence of one like ours needs no explanation.
In both cases, though, the dodge isn’t perfect. To really explain any given evil, Leibniz would have to show why it is secretly necessary in the face of a greater good (and Pangloss spends Candide trying to do exactly that). To explain any given law of physics, the multiverse needs to use anthropic reasoning: it needs to show that that law needs to be the way it is to support human-like life.
This sounds like a strict requirement, but in both cases it’s not actually so useful. Leibniz could (and Pangloss does) come up with an explanation for pretty much anything. The problem is that no-one actually knows which aspects of the universe are essential and which aren’t. Without a reliable way to describe the best of all possible worlds, we can’t actually test whether our world is one.
The same problem holds for anthropic reasoning. We don’t actually know what conditions are required to give rise to people like us. “People like us” is very vague, and dramatically different universes might still contain something that can perceive and observe. While it might seem that there are clear requirements, so far there hasn’t been enough for people to do very much with this type of reasoning.
However, for both Leibniz and most of the physicists who believe anthropic arguments, none of this really matters. That’s because the “best of all possible worlds” and “most anthropic of all possible worlds” aren’t really meant to be predictive theories. They’re meant to say that, once you are convinced of certain things, certain problems don’t matter anymore.
Leibniz, in particular, wasn’t trying to argue for the existence of his god. He began the argument convinced that a particular sort of god existed: one that was all-powerful and benevolent, and set in motion a deterministic universe bound by logic. His argument is meant to show that, if you believe in such a god, then the problem of evil can be ignored: no matter how bad the universe seems, it may still be the best possible world.
Similarly, the physicists convinced of the multiverse aren’t really getting there through naturalness. Rather, they’ve become convinced of a few key claims: that the universe is rapidly expanding, leading to a proliferating multiverse, and that the laws of physics in such a multiverse can vary from place to place, due to the huge landscape of possible laws of physics in string theory. If you already believe those things, then the naturalness problem can be ignored: we live in some randomly chosen part of the landscape hospitable to life, which can be anywhere it needs to be.
So despite their cheaty feel, both arguments are fine…provided you agree with their assumptions. Personally, I don’t agree with Leibniz. For the multiverse, I’m less sure. I’m not confident the universe expands fast enough to create a multiverse, I’m not even confident it’s speeding up its expansion now. I know there’s a lot of controversy about the math behind the string theory landscape, about whether the vast set of possible laws of physics are as consistent as they’re supposed to be…and of course, as anyone must admit, we don’t know whether string theory itself is true! I don’t think it’s impossible that the right argument comes around and convinces me of one or both claims, though. These kinds of arguments, “if assumptions, then conclusion” are the kind of thing that seems useless for a while…until someone convinces you of the conclusion, and they matter once again.
So in the end, despite the similarity, I’m not sure the multiverse deserves its own Candide. I’m not even sure Leibniz deserved Candide. But hopefully by understanding one, you can understand the other just a bit better.
The Niels Bohr Institute is hosting a conference this week on New Ideas in Cosmology. I’m no cosmologist, but it’s a pretty cool field, so as a local I’ve been sitting in on some of the talks. So far they’ve had a selection of really interesting speakers with quite a variety of interests, including a talk by Roger Penrose with his trademark hand-stippled drawings.
Including this old classic
One thing that has impressed me has been the “interdisciplinary” feel of the conference. By all rights this should be one “discipline”, cosmology. But in practice, each speaker came at the subject from a different direction. They all had a shared core of knowledge, common models of the universe they all compare to. But the knowledge they brought to the subject varied: some had deep knowledge of the mathematics of gravity, others worked with string theory, or particle physics, or numerical simulations. Each talk, aware of the varied audience, was a bit “colloquium-style“, introducing a framework before diving in to the latest research. Each speaker knew enough to talk to the others, but not so much that they couldn’t learn from them. It’s been unexpectedly refreshing, a real interdisciplinary conference done right.
The blog Astral Codex Ten had an interesting post a while back, about when to trust experts. Rather than thinking of some experts as “trustworthy” and some as “untrustworthy”, the post suggests an approach of “bounded distrust”. Even if an expert is biased or a news source sometimes lies, there are certain things you can still expect them to tell the truth about. If you are familiar enough with their work, you can get an idea of which kinds of claims you can trust and which you can’t, in a consistent and reliable way. Knowing how to do this is a skill, one you can learn to get better at.
In my corner of science, I can’t think of anyone who outright lies. Nonetheless, some claims are worth more trust than others. Sometimes experts have solid backing for what they say, direct experience that’s hard to contradict. Other times they’re speaking mostly from general impressions, and bias could easily creep in. Luckily, it’s not so hard to tell the difference. In this post, I’ll try to teach you how.
For an example, I’ll use something I saw at a conference last week. A speaker gave a talk describing the current state of cosmology: the new tools we have to map the early universe, and the challenges in using them to their full potential. After the talk, I remember her answering three questions. In each case, she seemed to know what she was talking about, but for different reasons. If she was contradicted by a different expert, I’d use these reasons to figure out which one to trust.
First, sometimes an expert gives what is an informed opinion, but just an informed opinion. As scientists, we are expected to know a fairly broad range of background behind our work, and be able to say something informed about it. We see overview talks and hear our colleagues’ takes, and get informed opinions about topics we otherwise don’t work on. This speaker fielded a question about quantum gravity, and her answer made it clear that the topic falls into this category for her. Her answer didn’t go into much detail, mentioning a few terms but no specific scientific results, and linked back in the end to a different question closer to her expertise. That’s generally how we speak on this kind of topic: vaguely enough to show what we know without overstepping.
The second question came from a different kind of knowledge, which I might call journal club knowledge. Many scientists have what are called “journal clubs”. We meet on a regular basis, read recent papers, and talk about them. The papers go beyond what we work on day-to-day, but not by that much, because the goal is to keep an eye open for future research topics. We read papers in close-by areas, watching for elements that could be useful, answers to questions we have or questions we know how to answer. The kind of “journal club knowledge” we have covers a fair amount of detail: these aren’t topics we are working on right now, but if we spent more time on it they could be. Here, the speaker answered a question about the Hubble tension, a discrepancy between two different ways of measuring the expansion of the universe. The way she answered focused on particular results: someone did X, there was a paper showing Y, this telescope is planning to measure Z. That kind of answer is a good way to tell that someone is answering from “journal club knowledge”. It’s clearly an area she could get involved in if she wanted to, one where she knows the important questions and which papers to read, with some of her own work close enough to the question to give an important advantage. But it was also clear that she hadn’t developed a full argument on one “side” or the other, and as such there are others I’d trust a bit more on that aspect of the question.
Finally, experts are the most trustworthy when we speak about our own work. In this speaker’s case, the questions about machine learning were where her expertise clearly shone through. Her answers there were detailed in a different way than her answers about the Hubble tension: not just papers, but personal experience. They were full of phrases like “I tried that, but it doesn’t work…” or “when we do this, we prefer to do it this way”. They also had the most technical terms of any of her answers, terms that clearly drew distinctions relevant to those who work in the field. In general, when an expert talks about what they do in their own work, and uses a lot of appropriate technical terms, you have especially good reason to trust them.
These cues can help a lot when evaluating experts. An expert who makes a generic claim, like “no evidence for X”, might not know as much as an expert who cites specific papers, and in turn they might not know as much as an expert who describes what they do in their own research. The cues aren’t perfect: one risk is that someone may be an expert on their own work, but that work may be irrelevant to the question you’re asking. But they help: rather than distrusting everyone, they help you towards “bounded distrust”, knowing which claims you can trust and which are riskier.
When most people think of the Big Bang, they imagine a single moment: a whole universe emerging from nothing. That’s not really how it worked, though. The Big Bang refers not to one event, but to a whole scientific theory. Using Einstein’s equations and some simplifying assumptions, we physicists can lay out a timeline for the universe’s earliest history. Different parts of this timeline have different evidence: some are meticulously tested, others we even expect to be wrong! It’s worth talking through this timeline and discussing what we know about each piece, and how we know it.
We can see surprisingly far back in time. As we look out into the universe, we see each star as it was when the light we see left it: longer ago the further the star is from us. Looking back, we see changes in the types of stars and galaxies: stars formed without the metals that later stars produced, galaxies made of those early stars. We see the universe become denser and hotter, until eventually we reach the last thing we can see: the cosmic microwave background, a faint light that fills our view in every direction. This light represents a change in the universe, the emergence of the first atoms. Before this, there were ions: free nuclei and electrons, forming a hot plasma. That plasma constantly emitted and absorbed light. As the universe cooled, the ions merged into atoms, and light was free to travel. Because of this, we cannot see back beyond this point. Our model gives detailed predictions for this curtain of light: its temperature, and even the ways it varies in intensity from place to place, which in turn let us hone our model further.
In principle, we could “see” a bit further. Light isn’t the only thing that travels freely through the universe. Neutrinos are almost massless, and pass through almost everything. Like the cosmic microwave background, the universe should have a cosmic neutrino background. This would come from much earlier, from an era when the universe was so dense that neutrinos regularly interacted with other matter. We haven’t detected this neutrino background yet, but future experiments might. Gravitational waves meanwhile, can also pass through almost any obstacle. There should be gravitational wave backgrounds as well, from a variety of eras in the early universe. Once again these haven’t been detected yet, but more powerful gravitational wave telescopes may yet see them.
We have indirect evidence a bit further back than we can see things directly. In the heat of the early universe the first protons and neutrons were merged via nuclear fusion, becoming the first atomic nuclei: isotopes of hydrogen, helium, and lithium. Our model lets us predict the proportions of these, how much helium and lithium per hydrogen atom. We can then compare this to the oldest stars we see, and see that the proportions are right. In this way, we know something about the universe from before we can “see” it.
We get surprised when we look at the universe on large scales, and compare widely separated regions. We find those regions are surprisingly similar, more than we would expect from randomness and the physics we know. Physicists have proposed different explanations for this. The most popular, cosmic inflation, suggests that the universe expanded very rapidly, accelerating so that a small region of similar matter was blown up much larger than the ordinary Big Bang model would have, projecting those similarities across the sky. While many think this proposal fits the data best, we still aren’t sure it’s the right one: there are alternate proposals, and it’s even controversial whether we should be surprised by the large-scale similarity in the first place.
We understand, in principle, how matter can come from “nothing”. This is sometimes presented as the most mysterious part of the Big Bang, the idea that matter could spontaneously emerge from an “empty” universe. But to a physicist, this isn’t very mysterious. Matter isn’t actually conserved, mass is just energy you haven’t met yet. Deep down, the universe is just a bunch of rippling quantum fields, with different ones more or less active at different times. Space-time itself is just another field, the gravitational field. When people say that in the Big Bang matter emerged from nothing, all they mean is that energy moved from the gravitational field to fields like the electron and quark, giving rise to particles. As we wind the model back, we can pretty well understand how this could happen.
If we extrapolate, winding Einstein’s equations back all the way, we reach a singularity: the whole universe, according to those equations, would have emerged from a single point, a time when everything was zero distance from everything else. This assumes, though, that Einstein’s equations keep working all the way back that far. That’s probably wrong, though. Einstein’s equations don’t include the effect of quantum mechanics, which should be much more important when the universe is at its hottest and densest. We don’t have a complete theory of quantum gravity yet (at least, not one that can model this), so we can’t be certain how to correct these equations. But in general, quantum theories tend to “fuzz out” singularities, spreading out a single point over a wider area. So it’s likely that the universe didn’t actually come from just a single point, and our various incomplete theories of quantum gravity tend to back this up.
So, starting from what we can see, we extrapolate back to what we can’t. We’re quite confident in some parts of the Big Bang theory: the emergence of the first galaxies, the first stars, the first atoms, and the first elements. Back far enough and things get more mysterious, we have proposals but no definite answers. And if you try to wind back up to the beginning, you find we still don’t have the right kind of theory to answer the question. That’s a task for the future.
I watched Hamilton on Disney+ recently. With GIFs and songs from the show all over social media for the last few years, there weren’t many surprises. One thing that nonetheless struck me was the focus on historical evidence. The musical Hamilton is based on Ron Chernow’s biography of Alexander Hamilton, and it preserves a surprising amount of the historian’s care for how we know what we know, hidden within the show’s other themes. From the refrain of “who tells your story”, to the importance of Eliza burning her letters with Hamilton (not just the emotional gesture but the “gap in the narrative” it created for historians), to the song “The Room Where It Happens” (which looked from GIFsets like it was about Burr’s desire for power, but is mostly about how much of history is hidden in conversations we can only partly reconstruct), the show keeps the puzzle of reasoning from incomplete evidence front-and-center.
Any time we try to reason about the past, we are faced with these kinds of questions. They don’t just apply to history, but to the so-called historical sciences as well, sciences that study the past. Instead of asking “who” told the story, such scientists must keep in mind “what” is telling the story. For example, paleontologists reason from fossils, and thus are limited by what does and doesn’t get preserved. As a result after a century of studying dinosaurs, only in the last twenty years did it become clear they had feathers.
Astronomy, too, is a historical science. Whenever astronomers look out at distant stars, they are looking at the past. And just like historians and paleontologists, they are limited by what evidence happened to be preserved, and what part of that evidence they can access.
Try to reason about the whole universe, and you end up asking similar questions. When we see the movement of “standard candle” stars, is that because the universe’s expansion is accelerating, or are the stars moving as a group?
For Halloween, this blog has a tradition of covering “thespookyside” of physics. This year, I’m bringing in a concept from biology to ask a spooky physics “what if?”
In the 1950’s, biologists discovered that birds were susceptible to a worryingly effective trick. By giving them artificial eggs larger and brighter than their actual babies, they found that the birds focused on the new eggs to the exclusion of their own. They couldn’t help trying to hatch the fake eggs, even if they were so large that they would fall off when they tried to sit on them. The effect, since observed in other species, became known as a supernormal stimulus, or superstimulus.
Can this happen to humans? Some think so. They worry about junk food we crave more than actual nutrients, or social media that eclipses our real relationships. Naturally, this idea inspires horror writers, who write about haunting music you can’t stop listening to, or holes in a wall that “fit” so well you’re compelled to climb in.
(And yes, it shows up in porn as well.)
But this is a physics blog, not a biology blog. What kind of superstimulus would work on physicists?
Abstruse goose knows what’s up
Well for one, this sounds a lot like some criticisms of string theory. Instead of a theory that just unifies some forces, why not unify all the forces? Instead of just learning some advanced mathematics, why not learn more, and more? And if you can’t be falsified by any experiment, well, all that would do is spoil the fun, right?
Do I actually think that string theory is a superstimulus, that astrophysics or particle physics is a superstimulus? In a word, no. Much as it might look that way from the news coverage, most physicists don’t work on these big, flashy questions. Far from being lured in by irresistible super-scale problems, most physicists work with tabletop experiments and useful materials. For those of us who do look up at the sky or down at the roots of the world, we do it not just because it’s compelling but because it has a good track record: physics wouldn’t exist if Newton hadn’t cared about the orbits of the planets. We study extremes because they advance our understanding of everything else, because they give us steam engines and transistors and change everyone’s lives for the better.
Then again, if I had fallen victim to a superstimulus, I’d say that anyway, right?
I don’t want to be the kind of science blogger who constantly complains about science fiction, but sometimes I can’t help myself.
When I blogged about zero-point energy a few weeks back, there was a particular book that set me off. Ian McDonald’s River of Gods depicts the interactions of human and AI agents in a fragmented 2047 India. One subplot deals with a power company pursuing zero-point energy, using an imagined completion of M theory called M* theory. This post contains spoilers for that subplot.
What frustrated me about River of Gods is that the physics in it almost makes sense. It isn’t just an excuse for magic, or a standard set of tropes. Even the name “M* theory” is extremely plausible, the sort of term that could get used for technical reasons in a few papers and get accidentally stuck as the name of our fundamental theory of nature. But because so much of the presentation makes sense, it’s actively frustrating when it doesn’t.
The problem is the role the landscape of M* theory plays in the story. The string theory (or M theory) landscape is the space of all consistent vacua, a list of every consistent “default” state the world could have. In the story, one of the AIs is trying to make a portal to somewhere else in the landscape, a world of pure code where AIs can live in peace without competing with humans.
The problem is that the landscape is not actually a real place in string theory. It’s a metaphorical mathematical space, a list organized by some handy coordinates. The other vacua, the other “default states”, aren’t places you can travel to, there just other ways the world could have been.
Ok, but what about the multiverse?
There are physicists out there who like to talk about multiple worlds. Some think they’re hypothetical, others argue they must exist. Sometimes they’ll talk about the string theory landscape. But to get a multiverse out of the string theory landscape, you need something else as well.
Two options for that “something else” exist. One is called eternal inflation, the other is the many-worlds interpretation of quantum mechanics. And neither lets you travel around the multiverse.
In eternal inflation, the universe is expanding faster and faster. It’s expanding so fast that, in most places, there isn’t enough time for anything complicated to form. Occasionally, though, due to quantum randomness, a small part of the universe expands a bit more slowly: slow enough for stars, planets, and maybe life. Each small part like that is its own little “Big Bang”, potentially with a different “default” state, a different vacuum from the string landscape. If eternal inflation is true then you can get multiple worlds, but they’re very far apart, and getting farther every second: not easy to visit.
The many-worlds interpretation is a way to think about quantum mechanics. One way to think about quantum mechanics is to say that quantum states are undetermined until you measure them: a particle could be spinning left or right, Schrödinger’s cat could be alive or dead, and only when measured is their state certain. The many-worlds interpretation offers a different way: by doing away with measurement, it instead keeps the universe in the initial “undetermined” state. The universe only looks determined to us because of our place in it: our states become entangled with those of particles and cats, so that our experiences only correspond to one determined outcome, the “cat alive branch” or the “cat dead branch”. Combine this with the string landscape, and our universe might have split into different “branches” for each possible stable state, each possible vacuum. But you can’t travel to those places, your experiences are still “just on one branch”. If they weren’t, many-worlds wouldn’t be an interpretation, it would just be obviously wrong.
In River of Gods, the AI manipulates a power company into using a particle accelerator to make a bubble of a different vacuum in the landscape. Surprisingly, that isn’t impossible. Making a bubble like that is a bit like what the Large Hadron Collider does, but on a much larger scale. When the Large Hadron Collider detected a Higgs boson, it had created a small ripple in the Higgs field, a small deviation from its default state. One could imagine a bigger ripple doing more: with vastly more energy, maybe you could force the Higgs all the way to a different default, a new vacuum in its landscape of possibilities.
Doing that doesn’t create a portal to another world, though. It destroys our world.
That bubble of a different vacuum isn’t another branch of quantum many-worlds, and it isn’t a far-off big bang from eternal inflation. It’s a part of our own universe, one with a different “default state” where the particles we’re made of can’t exist. And typically, a bubble like that spreads at the speed of light.
In the story, they have a way to stabilize the bubble, stop it from growing or shrinking. That’s at least vaguely believable. But it means that their “portal to another world” is just a little bubble in the middle of a big expensive device. Maybe the AI can live there happily…until the humans pull the plug.
Or maybe they can’t stabilize it, and the bubble spreads and spreads at the speed of light destroying everything. That would certainly be another way for the AI to live without human interference. It’s a bit less peaceful than advertised, though.
4 gravitons 