Physicists Criticize Stephen Wolfram’s ‘Theory of Everything’
scientificamerican.com
scientificamerican.com
There are lots of people who dislike him, great, I agree, this is common knowledge, but it's not a reason that his research is wrong – the scientific community is better than ad-hominem attacks.
The problem is that he's a pretty good science communicator. As a layperson (but software engineer so I'm somewhat mathematically literate and know what a graph is), I got through the intro talk published by them and came out feeling like I understood what the premise was, roughly how it worked, how it could apply, why it might be a good model.
I then read a physicists "debunking", but because it was basically just references to physics papers saying "ah but you didn't address the something conjecture" I didn't really get anywhere with it.
I'm glad the physics community are discussing it internally, that is ultimately more important, but as a layperson I'd really like to get a good understanding of the other side of the argument.
This is like saying you've invented a car from scratch because you've made something with mirror symmetry that can roll forward. Obviously, this is missing more than a few details.
What Stephen Wolfram seems to have done is come up with an extremely generic system that can describe anything, in much the same way that a CAD program can use meshes to describe any piece of machinery you'd like to design. That doesn't mean that meshes are equivalent to metal and concrete. Just because meshes can have various symmetries and can be used to represent circular things doesn't mean that a mesh is a car either.
Similarly, in a very real sense the Universe performs computation, because -- duh -- it contains computers. And brains. And many things of that nature. Brains and computers run by the same rules of physics that stars and experimental particle beams do. Hence, the Universe computes. It also computes in the trivial sense that if you add 1 hydrogen atom to 1 hydrogen atom you get 2 hydrogen atoms.
So if you come up with a very generic form of computation, such as automata, or substitution rules, or whatever, of course you're going to find many things that "look like" various aspects of the computation the Universe is performing. But analogies are just that: comparisons of vaguely similar features. That doesn't mean they're isomorphic to each other, nor does it mean that some random theory is actually the rules the Universe follows.
Stephen spends a lot of his time jumping up and down and pointing at some trivial aspect of his theories and saying "Look at this! My theory can perform addition! The Universe does addition too! Hence my theory is the theory of the Universe!"
I've been interested in the fundamental theory of everything for two decades now, and I spend about an hour a day on average poking through random arXiv papers looking for ideas or signs of progress.
I've noticed a few things: Much of Wolfram's criticism of academia and modern theoretical physics is correct. It's much the same as Eric Weinstein's criticism. They both say that the politics and economics of academia make it nigh impossible to rock the boat and come up with radical new ideas. I noticed this too. Modern theoretical physics isn't. It's intellectual wankery and dick measuring as a way of getting pieces of paper and tenure. There's zero motivation or incentive to do anything useful, because the system has slowly evolved over time to disincentivize it.
However, this doesn't mean that their particular "rebel" flavour of TOE is correct, just because the models produced by Universities have gone nowhere. Progress in some arbitrary direction is not necessarily better than no progress at all! Progress in the right direction is needed.
Over the years (decades!) I've developed an acid test to quickly sort the wheat from the chaff:
1) Does the theory even mention particle generations? Bonus points if demonstrates at least three and explains why we haven't found four.
2) Is it a truly local theory in that global communication is never required? (Many current theories implicitly involve "solving" something, which is often implicitly a global minimisation.)
3) Can the theory be applied to single systems, or just ensembles? (Most QM theories automatically fail this one. Single particles exist, single electrons go into higher orbits from a single photon, etc... Either you can model this, or your theory is woefully incomplete)
4) Can it explain the observations of Quantum Mechanics without resorting to the "infinite dimensional" mathematical wizardry? Can it produce computable models, or just symbolic gibberish that I guarantee you no particle "computes" as it goes about its business.
5) Nobel prize winning stuff would explain the origins of constants that are currently experimentally determined inputs. (Most theories I've seen don't even have numbers.)
6) The ultimate requirement is for a TOE to "not only" produce the known observable physical behaviour but to also naturally exclude everything else. That is, the ideal rules should have some sort of incredibly tight self-consistency that can only happen in one way and can only produce physics. For example, an algebra that can only be used to represent valid combinations of particles and fields, and can never represent any other "invalid state" that makes no sense.
Unfortunately, 99% of the stuff I see fail at all of the above. Wolfram's theory does too. His current graph-based theory and his previous cellular automaton theory spectacularly fail the 6-th requirement. They're both nearly unconstrained computation models with infinite flexibility, the exact opposite of a good TOE.
I'm interested in discussions about why this type of model can't work, rather than people repeatedly pointing out that they haven't found the exact model yet, which seems to be a strawman argument.
As for your 6 acid test elements, that's exactly the sort of thing I'd like to learn more about, but in the current wording I feel like I'm missing about a Phd's worth of understanding. I think communicating this in a more accessible form would be great!
I understand that theoretical physicists have reasons why they don't like his theories, but I don't understand what those reasons are, whereas I do understand a bit about the theories. This makes it hard (as much as I want to) to be persuaded by physicists.
That's not even wrong! Studying computation, computation systems, or even just mathematics in general almost certainly could assist with progress towards a TOE.
That doesn't mean that pointing at a computer running a few dozen lines of code[1] and saying "A TOE could be in there... somewhere" is materially useful.
[1] I'm paraphrasing Stephen himself here!
Personally, I feel that if Stephen really wanted forward progress towards a TOE, he's uniquely positioned to do so. He controls a company that makes one of the best Computer Algebra Systems. He's in a position of power over a large, highly talented staff of mathematicians, and he could push Mathematica's features towards multiple relatively unexplored areas of computational mathematics where there isn't much available at the moment.
Unfortunately, he's made the decision that the only two such areas worthy of his time and his company's resources are cellular automata and graph rewriting.
Sigh.
Meanwhile, he could have done added all sorts of features for things like handling Einstein summation convention, tensors, Geometric Algebra (and noncommutative algebras in general), topology, etc...
Those would at least have immediate utility to working physicists, and if pushed to the extreme could maybe help uncover some fundamental theory somewhere.
But nope. He likes to play with graphs, like a kid with with a Spirograph only interested in making pretty patterns.
Are python libraries taking up this challenge?
The parent articulated that this is psuedoscience, preached by a man who has advocated self-serving psuedoscience before.
Again, as I've said in other comments, just pointing at him and saying he's a fraud isn't useful in explaining why. I think the scientific discourse deserves better than that.
It is also uncomfortably close to the tactics of creationists, antivaxxers, and other conspiracy theorists.
I've done science communication myself, I helped run a science festival, I know many people who still do it and are very good at it. I know how important it is to communicate well and how much impact it can have! I've even seen good science communication turn creationists and antivaxxers.
I must admit I had seen the parallels while writing these things, but I've also seen people who unfortunately do believe these things turned away from mainstream science because they aren't yet at a level of understanding that the community mandates they operate at.
Unfortunately for the mainstream scientific community there's often a disdain for communication like this. The fact that Wolfram is relatively good at it makes the whole situation more difficult.
Economics is also full of theories that try to explain the entire complex system from partial observations, and people arguing furiously with each other, in what for a large part in academia is a contrived nonsense that is practically useless for real world issues, and mainly serves signalling purposes.
How this manifests itself there, is that they will take some time series of some kind of data - price of oil over a few years say - fit the equation of a straight line to it - and the less responsible ones then yell "solved". But if you took a longer portion of that data, went back several decades say, you would see that it is bouncing up and down like a yo-yo, and actually the only equation that is pretty much guaranteed not to fit it, is a straight line. Unless you take a short enough period. Or to put it another way, with a short enough ruler you can fit any curve.
Wolfram, and his opponents, are essentially doing the same thing, only with multi-dimensional math, where the curve in question looks a bit more complex than a straight line, but fundamentally its the same thing. Which is why it matters much more than outsiders may appreciate that the model isn't an exact fit.
Lurking behind all of this of course, is that one time in physics when the data and the theory were (within the limits of then technology) a pretty exact fit, and still the theory was completely wrong - epicycles.
In other words...they haven't made any significant discovery, yet.
The physicists are just saying "OK, get back to us when you have something concrete we can evaluate."
'Does your startup implement every single feature of Windows / Excel / Postgres / AWS and then some? No? Well, then go back to work and come back when you have something truly innovative.'
There is no charity. 'What you did is fairly incomplete, but the ground you covered so far is promising. Here's a list of challenges. Where would you start addressing them?' Maybe the answer will come 10 years down the road, in some University unaffiliated with the original author. In the current system, there is zero investigation of anything else but well trodden orthodoxy.
When ZFS was announced, it blew my mind. The Sun ZFS team showed that traditional RAID had a veritable laundry list of fatal, nigh unfixable flaws. Various forms of silent data corruption caused by things such as reordered writes, partial writes, lost writes, or bit flips, and the like.
Then they showed how Merkle Trees provide a simple, unified, provably correct mathematical model of data integrity that in one fell swoop permanently eliminates all of the issues with traditional RAID data integrity. Poof. Gone! Reordered writes: detected. Lost writes: detected. Bit flips: detected. ZFS could even protect against malware in the drive firmware. RAID has no hope.
They demonstrated that in their testing, they had uncovered bugs in several commercial RAID adapter cards and even the DMA chip of one particular motherboard. These were real problems that really were eating people's data!
Did they have online ZFS pool expand on day 1? No. Defrag? No. Encryption? Nope. In fact, it took over a decade to add all of the various features and niceties that people wanted.
That doesn't matter though, because on Day 1 they demonstrated a fundamentally better way of storing data.
Stephen Wolfram hasn't really done this. To put things in the above terms, he hasn't managed to even store a file and retrieve it intact yet, and there's no firm indication that this will ever be possible. This makes his system indistinguishable from /dev/null.
Worse, his system may be the type where you can put files in, but when you try to read them back out, the filesystem simply contains an infinite list of files, one for all possible bit strings. I mean sure, technically your file is in there... somewhere. But this is indistinguishable from /dev/random. Also not useful.
1. The model must be fundamentally simple.
2. The model must be able to produce 'interesting' behaviors somehow.
3. The model must exhibit very puzzling behaviors like 'speed of light is a constant', 'space dilation' or 'quantum superposition'.
He observes that extremely simple rewrite systems, like the infamous rule 30, satisfy 1 and 2, under a provable universality result: all 'interesting' behaviors are Turing equivalent. He then observes that the succession of states of the system could be observed as 'time', that having the fundamental simulation substrate create new nodes on the fly could be observed as 'space dilation', that the speed of generating new nodes is a limiting factor in the same way the speed of light is and that having alternate rules applicable in a given configurations could be observed as 'superposition'. This sounds promising. Let's work out the details. To my layman understanding, it's a fundamentally different approach than what physicists have been doing, extremely successfully, in the past 100+ years.
So far, he's empty handed, and very far from concrete results. He's pioneering the exploration of this space by building the first generation of programming languages, the machine code. Maybe 50 later, someone will build ZFS / explain the '3 but not 4 particle generations', using ML to express their thought process.
He's put some thought in attempting to bridge graph rewrite systems to the Universe's behavior. No measurable results, but no [layman's] obvious way to refute either. https://www.wolframphysics.org/technical-introduction/potent...
But he wasn't.
I am not saying he hasn't done anything interesting, but it is actually hard to tell exactly where any original work begins, even for someone somewhat familiar with the area of computing related to graph transforms (me), because in his writing he implies he did so much more than he did. So it isn't clear where his contribution line starts.
It rubs me so wrong I can't read anything he writes anymore.
This may be one aspect of the negative reaction of some physicists too.
While I'm not familiar with what kind of graph transforms work is done these days, I would have hoped that if someone re-created some version of Lorenz transforms on top of graph transforms, I would have heard of it. The discussion would be more constructive, or destructive as a manner of speaking, if someone would point out that either Stephen is wrong, or old news, on the handful of technical points he is making.
https://www.wolframphysics.org/technical-introduction/potent...
I wish that you had linked to the material that you didn't "get anywhere with"; without it, it's hard to actually examine the quality of the underlying arguments. I found writeups like [0] which seem extremely well-argued; the formal ideas come from [1], and the core idea is Bell's theorem [2]: there is some essential non-locality to QM, we can experimentally see it, and so we can't just say "oh, it's all a big computer", even if it's a computer with interestingly high dimensionality or interconnections, without further justification.
[0] https://www.singlelunch.com/2020/04/23/why-stephen-wolframs-...
My understanding is that they are early on in the research so the lack of explanatory power feels... uninteresting at this stage? They have started to make predictions (such as the existence of lighter weight particles than electrons that still have mass), but in general it just feels like they're still too early to conclude that it's not falsifiable and has no explanatory power?
That's why it's important to point out that it's not falsifiable and has no explanatory power at the moment. What he's doing is bad for science and does real harm, eroding the public understanding of how science actually works in favor of flashy headlines.
If you read/listen/watch the announcements (which I'll agree are a bit over the top), the position is very clear that they have "found" a new avenue of research that they believe has the possibility to bear fruit. They are clear that it _could_ form the basis for a "theory of everything", once much more research has been done.
They are optimistic in how they talk about it, and many may disagree with that, but they are not claiming they have "solved physics".
It often can be quick to determine whether a theory will ever be falsifiable - it has to do with the nature of the claims being made by the theory.
> such as the existence of lighter weight particles than electrons that still have mass
That's not a prediction, as we know that neutrinos exist.
String theory also predicts some things that are congruent with existing theories but, by dint of building them differently, gives different predictions for how they work. In this, we get new explanatory power. For example, string theory predicts fuzzball black holes, giving a resolution to the firewall problem. All string theories have tachyons, which is surprising but would not be the first time that they've shown up theoretically, and they solve some questions about spacetime (while opening many others!) Some string theories have photons and gravitons, which can unify our existing conflicts between GR and QM.
And, crucially, string theory admits narrowing of conjectures and explanatory regimes. It's easy to take a string theory, combine it with an observation from a satellite or a particle accelerator, and conclude that no, the theory was wrong. Horribly wrong, completely wrong, totally wrong. String theory is actually sufficiently formal that the current source of realizing wrongness is from "swampland" critiques, which are merely observations that a working string theory ought to pass some basic arithmetic sanity checks.
Sure, maybe my conclusions are premature. As a cousin comment notes, though, a particle theory that doesn't immediately spit out the Standard Model is a little suspect, and one that admits basically any model of particle physics is also a little suspect.
In other words, it's not science yet.
Disagree. I bought one of his books, Idea Makers, knowing that he is a bit a jerk/kooky, because it looked interesting. Every chapter about a different important scientist/thinker in history, what could go wrong? What went wrong was that actually every chapter was about Wolfram and/or Mathematica. He can't help himself!
But, all that said, that doesn't mean he is wrong!
Seems like a pretty smart guy, though, and has built up a pretty impressive empire.
But my bet*) is that if it gets accepted, then with severely reduced claims (not "theory of everything", but "We model X with Y, and were able to replicate Z1 in general and Z2 in low-energy limit. Further studies...").
Large company and many people involved, as of now, is the only reason why it gets some attention. Otherwise it would be buried as a PDF in "misc" section on one's academic website.
I'm a physics postdoc and can't understand what he writes. How good of a communicator can he possibly be?
He's pretty good at sounding like a good science communicator. For someone without a Phd in physics, this is unfortunately indistinguishable from actually being a good science communicator.
The first part without the second one could mean "great communicator". The first part with the second one sounds more like "crackpot".
It is easy to impress people when you have studied something longer than them. There are interesting ideas that are relatively simple to learn, so if you learn them and then explain them simply to other people, they are amazed.
But it is possible to get some beginners' stuff right, and then get some advanced stuff wrong. And the laypeople who listen to you are impressed by the former, and cannot notice the flaws in the latter. Then you claim to have found answers to some expert-level questions, and your fans go like "this is pure genius, why doesn't academia build statues in his honor?"
From my perspective, the part that impresses people is Wolfram pointing at a Turing-complete system (it used to be cellular automata, recently it is some graph transformations) and explaining what Turing-complete means. Which is fantastic if you never heard about Turing-complete systems before! Yes, in theory it could be used to simulate a universe, despite being such a simple thing. If this is news for you, it will completely blow your mind!
On the other hand, if you studied computer science, you know that Turing-complete systems are a dime a dozen. For example, almost any programming language is Turing-complete. Yes, it is fascinating to find things that are way simpler than the usual programming language, and are still Turing-complete. But again, if you studied computer science, you already heard about a few of them, so learning that there is yet another one doesn't sound so revolutionary. At best, you may admire its elegance, but it's still only one specimen amongst many.
So when Wolfram continues to explain how his currently favorite system could simulate the universe, you start yawning, because "being able in principle to simulate the universe as we know it" is a general property of Turing-complete systems. Yes, it is a property of Wolfram's current favorite system, but it is also a property of zillion other Turing-complete systems. There is no reason to single out this one specifically. You might as well talk about how the universe could in theory be simulated in PHP. Yes, it could, but without further evidence there is no reason to assume it really is.
(In fact, it's even worse. If something is Turing-complete, not only it could simulate our universe, but also any other universe with completely different laws of physics. Therefore, despite the fact that our universe could be simulated in PHP, you can't really hope to discover anything relevant about our universe by studying PHP. A universe without gravity, or a universe where gravity is a repulsive force, could be simulated in PHP, too. This is why you cannot find useful information about gravity by studying what can be done in PHP.)
...so far, this was inside my field of expertise, which is computer science; now I am going to step outside of it and comment on physics...
If I understand it correctly (and reading comments by people who know physics better than me suggest that I am roughly on the right track), the next step is Wolfram finding some vague analogies between some properties of his currently favorite Turing-complete systems, and the laws of physics as we know them. Like, if something grows, that is kinda like the expansion of universe; but if something shrinks, that is kinda like the relativistic contraction; and if two things are somehow connected to each other, that is kinda like the quantum entanglement. Maybe the actual analogies found by Wolfram are less lame than this, but ultimately they are just analogies, and probably sound equally lame to an expert in physics. Things kinda grow, but not in a way that would allow you to actually derive the equation for Lorentz contraction. They just grow, and that's it. If we wouldn't have the theory of relativity by now, you couldn't use this to invent it. It has no predictive power. First you must know the laws of physics, and only then you can search for something that is vaguely analogical to them.
And it is this combined effect of "finding a simple system that is Turing-complete" and "finding some analogies between the system and the laws of physics" that creates the strong impression in laymen that Wolfram found something relevant to understanding our universe. "Hey, if it can be used to simulate the universe, and if we can find analogies with some laws of physics, why don't we just study it further and maybe we will find some new, hitherto unknown, laws of physics? Why is the academia not paying more attention?" The main problem is that the fact that it could simulate the universe actually means much less than how it sounds, and the analogies can only be found when you know what you are looking for, in other words only after the relevant laws of physics have been already discovered.
EDIT:
Looking at this discussion, you can find people who defend Wolfram because he sounds exciting, or because it feels unfair that his theories do not get the attention they supposedly deserve. Not a single person is defending him because he explained some law of physics correctly. And this is not an accident. As far as I know, the same thing happens outside Hacker News, too.
That includes variability in some things that are constant in our region of the universe. It's fine for everyday purposes for us to regard things like the ratio of masses of the particles or strengths of the forces as fundamental and given, but a truly fundamental theory needs to generate the entire landscape, possibly including variation in the number of extended dimensions of space.
At the same time, there is a sense in which it's easy to do that: just use a universal Turing machine. A correct fundamental theory should generate something akin to the landscape of string theory, but at the same time not generate e.g. Conway's Game of Life. Is Wolfram's theory being evaluated on achieving that delicate balance between generating too little and too much?
Wolfram published "New Kind of Science" almost 20 years ago. Scott Aaronson was kind enough to review it: https://scottaaronson.com/papers/nks.pdf So was Juergen Schmidhuber and many others http://shell.cas.usf.edu/~wclark/ANKOS_reviews.html
Wolfram has now moved from CA to Hypergraphs (essentially sets). They are so generic structures that you get anything physics-like from them if you want.
For example, if you visualize hypergraph that you get from normal execution of numerical simulations for relativistic, quantum or general relativity simulations, they would look cool and you might realize that they are connected to physics. Just like Wolfram's pretty pictures. Would execution graph representation explain more than equations and code? I don't think so.
At the same time it misses the point of the natural sciences.
Hypergraphs as described are more like Calculus than Physics. Of course, describing gravity with calculus was a brilliant discovery but you can't just have calculus and say "that's gravity". You need the connection between the two.
His model is inherently computational. It would be like saying "here's a computer that can execute things" and then proving that you'd created the Linux kernel or something.
It's not a very surprising claim to say that a particular fact can be represented in a system which can represent anything. Mathematicians (well, most anyways) don't go around claiming they've discovered the secrets of the universe because math _could_ describe it in principle.
I think he's excited because he thinks that in this framework we can just "throw compute" at the problem to enumerate the possible theories of everything, and automatically rule out the ones that don't align with reality.
I've often wondered why they don't get the same vitriol. I think it's mainly because they're still attached to physical formalisms: gauge theories, algebras, known symmetries and conservation laws, etc. It at least has some constraints, even if they're not testable.
Back 20 years ago when Wolfram released A New Kind of Science, my GR teacher said about that book: "Science makes predictions, not press statements." Even Hawking got in a lot of trouble in his later years when he'd announce all those extreme claims and walk them down without ever publishing anything. It really surprises me how much press coverage this guy can get over 2 decades without actually making a theoretically testable prediction.
I'm not sure if this is because of the hyper competitive state of getting tenure or if this is just an example of group think against the other, but I don't think it's a coincidence that the field of physics has gone from the most exciting and impactful field in science to being largely stagnant.
"Proof? Who needs that! My theory is so elegant!"
"What do you mean my proof doesn't make any sense, maybe I'm just too much of a genius! How can you not see that right away?"
"You have, 1028 emails with frustratingly naive 'What About?' questions from various self-taught physicists."
I think a lot of physicists get kinda jaded about outside work. Then, when a genuinely talented and hardworking non-professional shows up, it's hard to tell the difference.
The lower hanging fruit has already been picked.
Whether that’s true or not the way he’s so quickly and casually dismissed people like that, that have made relevant contributions to his work is a red flag for me.
P.s. normally Conway’s work wouldn’t overlap centrally with a theory of everything but it’s more relevant given Wolfram’s approach.
"An interferring CEO who thinks he knows everything and any project can completed by one engineer in 2 weeks, maybe 3 if it's really complicated
[...]
If you do have to work with Stephen Wolfram, may god help you. 30 years ago, he was a genius with ideas ahead of his time and he managed to productize them. Today, he is still a genius with ideas that would have been ahead of their time 20 years ago. He is failing to lead this company to the future, and his main method of leading is screaming like a toddler who is told they can't go to the beach during a hurricane.
If you are in upper management or project management, you will have to meet with him multiple times per week. In these meetings he will ask for updates on features and products he invented a week ago and be angry they are not ready to be pushed to production yet. After he finishes loudly complaining about how slow and bad everyone is, then he'll go on to request entirely new things that he expects to get immediately. Priorities are never clear, and when you try to get him to guide you, he he has no idea what he actually wants other than "everything".
His irrationality and personal attacks drive people to tears."
Maybe I am overly harsh, but to me, it sounds like Steve Jobs minus being actually at the frontier of innovation.
Wolfram just sounds like a very intelligent but childish and insensitive leader with bursts of anger because reality doesn't follow his way.
Unless you are going to work for him, it is largely irrelevant whether he's a megalomaniac or not. He's a person that has produced very interesting and useful work (e.g., mathematica). You can take the good things about him and forget about his pompous/annoying stuff.
A plausible explanation for his being viewed as megalomaniac is a disapproval of a perceived lack of humbleness ("A good scientist ought to downplay their discoveries, make them seem like no big deal. He claims his discoveries are super important, so he's a baaaad guy").
Having been in your (supposed) position of coming across discussions about Wolfram, where the first things I read where judgments about Wolfram the person instead of his ideas, I arrived at the following conclusion after a while:
I don't know and don't have the expertise to evaluate the correctness or value of his ideas to physics or science in general. I'm also not interested in "drama", I just want to learn cool stuff. So what I can do is actually read his work, keep an open mind with respect to his writing style, and see what, if anything, I get out of it.
The result of that is that I occasionally read a blog post by him, or a random paragraph from NKS (which is available online), or watch one of his Q&A streams. [2] And from those, I get tremendous value. The ideas are often beautiful, clearly explained and visualized, and some have had a large impact of my thinking about both the universe and the mind.
Not reading someone's work because of other peoples' judgment of that person just robs you of potential opportunities to learn something.
[1] https://people.well.com/user/abs/Cyb/archive/c3m_0201.html
[2] https://www.twitch.tv/stephen_wolfram/videos?filter=archives...
Edit: Downvoters, got evidence? Here's Gorard's papers [0] [1], and Wolfram is clearly taking credit for this work upon which only Gorard's name appears. Gorard doesn't do themselves any favors, by name-dropping Wolfram constantly. Almost as if Wolfram employs Gorard and pressures them heavily to incorporate the name-dropping...
[0] https://arxiv.org/abs/2004.14810
[1] https://www.wolframcloud.com/obj/wolframphysics/Documents/so...
Remember: no-one ever claimed the prize for disproving the TimeCube guy either.
And when provided with some of the responses from other physicists regarding his work, Wolfram is singularly unenthused. “I’m disappointed by the naivete of the questions that you’re communicating,” he grumbles. “I deserve better.”
And yet Einstein did just that in developing General Relativity, as have many other scientists and thinkers in producing their novel theories or approaches. Yes, GR made falsifiable predictions, but it still represents a paradigm shift, the sort Katie Mack thinks of as "unfortunate".
I am not endorsing Wolfram's work at all. But the idea that science has to progress in a step by step, gradual, collaborative fashion is naive and historically ignorant.
Einstein was an obvious outlier, and Wolfram is no Einstein.
Your argument is used by every crackpot to justify their dismissal of the scientific process - because making progress that way is much harder than simply posting their theory on a primadonna blog and demanding the world come to them.
We need both sorts of thinkers in science: those who want to make the important incremental improvements and those who seek entirely new ways of looking at problems.
It's also important to not immediately discount Wolfram's work simply because he is wealthy and successful and not a formal member of academia. However it seems that some are personally threatened or dislike Wolfram because of who he is which triggers a knee jerk reaction against his work.
I am not an internet fanboy of Wolfram's, I'm only pushing back a bit against the idea that incremental progress upon the scientific mainstream is the only path forward.
I and others are just suggesting Wolfram go through the same process as everybody else instead of using his power to subvert it. That isn't motivated by being threatened, just by a sense of fairness.
As for his good ideas, absolutely - but they are all in business and unconnected to the grand theories he is now promoting - his prior physics experience is junior and long ago, and certainly doesn't warrant special treatment.
Einstein was no Einstein either until he became Einstein. You shouldn´t dismiss a paradigm shifting idea just because those are rare. Otherwise, you might miss out on the real one when it comes along. Instead, examine the ideas presented critically and see where they lead you. nothing wrong with that.
As for the crackpot part, science is filled with instances and examples of things that were considered to be crackpot ideas that proved to be not only true, but completely world changing. Have a little bit more humility please rather than being dismissive. Not saying that he´s right but it´s worth listening to the arguments made in case there is some truth to be found in them.
The humility required is from Wolfram, and if he put his arguments in the appropriate forum, they would be listened to, but they would have to fight it out with all the others.
It's certainly not worth listening to every pet theory posted online. They are ten a penny.
This does not intend to take any little bit of merit from Einstein's gigantic contributions, it just come to show the point that Katie Mack is making about the role of community driven refinement. Nordström built on Special Relativity and Einstein's ideas to build a geometric theory of gravity, this one turned out to be incorrect as it didn't predict the correct amount of light bending or perihelion precession. But some slight changes produces the correct one. Nowadays we know that Nordström's theory got only one thing wrong, it's a scalar theory of gravity. Change the spin of the gravitational field to spin 2 and you get Einstein's theory.
Wolfram too is refining the work of others as mentioned in the article.
The fact that someone else came up with the EFE at essentially the same time as Einstein is telling.
This doesn't change the fact that Einstein's approach was a radical departure from the prevailing mainstream. Just as Wolfram's is today.
Trying to classify Einstein as an incrementalist on the basis of his work being built upon previous work would then imply that Wolfram is also an incrementalist because he too is inspired by earlier work in the same vein as his, as the article states. But clearly we're talking about something different here.
But from most simple observations in nature, there's no direct causal link between the two, . Intuitively, we "know" everything originated from a "Big Bang", but we don't have proof how or if complexity arose over time, or was it always intrinsinct to cosmos, and wether the question itself have any measurable implications..
So graphs and computational feedback remain interesting and time-consuming avenues for research, but it's a big leap from superficial studies of outputs ("how") and finding a graph-formula that would explain all possible universes ie. from a "Big Bang" point ("why").
A solid unchanging theory would need to address enough "why"'s in order to be properly backed up by predictions explaining the observable physical universe, to be relevant. If the theory can include new hypothesis and rules at any time, it's not a theory, but a container for information and rulesets (ie. a Turing-complete system).
Of course, you can "curve fit" anything inside a graph. They're great at it for simulations and games, and are already used everywhere you do computationally interesting complex stuff. There's just no indication or proof that the physical world matches any model created by humans living within the universe. Maybe someone humble will stumble unto some simple explanation, but so far, the deeper we go, it seems even experts get lost in the tough complexities and unintuitive laws we assume govern microcosmos and energy.
It seems to boil down to difference between theory and practice. Pure mathematics and observations are theory, while how to simulate and mimic, is more engineering and providing interesting outputs. If we combine the two, we can build universes.. (ok, off the pipe for now :)
I don't think we have any intuition about this at all, we are however fairly accustomed to this idea with old creation myths that still loudly resonates throughout culture, and of course now that it has been the scientific orthodoxy for 50 years or so.
But yeah I definitly agree with you on the "curve fit" thing. I would not be surprised that, if one day we do find a comprehensive theory of everything (assuming this is even possible or even have a meaning at all), it would be possible to express it in this very general framework of hyper-graph re-writing itself according to some rule. That doesn't mean it's a good idea to so.
And in any case, at the end of the day, like everything else humans do, science is a social process. The goal is to get closer to truth (whatever that means), but you won't get there by ignoring any input, and insist that everyone should drop whatever they are doing and come admire/help you on your work.
CA does not seem to be great at replicating conservation principles -- and they are everywhere in physics -- and attempts to graft them on by complex interactions between glider guns or the like only highlights the fragility of the process, as even the slightest misalignment causes chaos. When you run a CA, you get expansion or contraction and are typically left with some stable/oscillating phenomena if it doesn't just keep blowing outward, and those configurations which remain are still ... fragile. In ways particles are not. Particles might come and go, but the quantities they represent linger. Where is that in CA?
As to anyone who suggests they have new physics, bring us testable observations of that which would not be explicable in our current frameworks. Give us a refinement of the numbers, measurable in some fashion, to highlight a deviation from that which is predicted by the old ways. I haven't seen anything like that out of him.
Perhaps a ruleset exists which would have these things, but if so, wouldn't he have mentioned it instead of just doing his usual business about Rule 30 and the like? Or are we -- and here is where, outside of ego, I get the biggest crank feelings from Wolfram -- all expected to drop everything, adopt his approach, and iterate through ever more complex configurations of CA until we find the one that somehow works? Of all of the bits in Baez's crackpot index, this is the one reliable indicator I find missing, although "10 points for each statement along the lines of 'I'm not good at math, but my theory is conceptually right, so all I need is for someone to express it in terms of equations'." is in the ballpark. Everyone, drop everything and do it my way until it works!
Just gotta keep adding more epicycles, eventually geocentrism will be vindicated! /s
I think it's a cool idea that may have results in the future, which is how I felt he presented it.
That has to be one of funniest quotes I ever read.
I did have this thought that Stephen Wolfram may be overrated in the past. He is a good businessman for sure. The amount of attention he gets from the media does appear to be disproportionate to his achievements and perhaps may cause overconfidence.
That said I do understand the view that the scientific community may have become so narrowly focused on rigid scientific review processes that they could be missing out on real innovation and alternative ideas.
For example I can't help but think of what would happen if some random (not PhD) engineer came up with a new and highly innovative mathematical construct and published it on their blog... I doubt their idea would get any attention at all; no matter how useful or impressive it is. It would be lost in the noise because scientists are not going to review it.
Science is still operating on an outdated idea that all scientific knowledge is centralized within universities and that all credible innovations must necessarily come from these universities and all great scientists must be somehow be socially connected to existing great scientists. As if direct social connections are still the only pathways through which cutting-edge knowledge can flow. It completely ignores the existence of the internet.
My field was an uninteresting sub-field of programming language theory, and every conference, at least 1/2 the papers submitted were borderline batshit insane. The stuff on the internet is gibberish.
Screaming into the night & railing that physicists are ignoring you is just sociopathy. Instead, go to them, give presentations, prove your ideas have merit. Make it short to get their attention, build up credit & lead them to what’s new.
Physicists by trade do not trade on advancing the state of the art. That takes courage and risk of reputation. This is why the natural sciences have progressed outside of vocations. “Lead then to what’s new” is better worded as “make it shameful to ignore you”. I am acutely aware of that now.
Anecdote: as an undergrad I worked part-time for a physics research group, and once my address got added to their staff web page, I started receiving "highly innovative" new theories via email about once per week. They were generally long and involved; it would have taken me hours to pore through one and figure out what it was trying to argue or whether it made any sense.
And this was circa 1995, when only a small portion of the world had access to email, let alone web pages. I wonder what it's like now...
There's no problem admitting he's forgotten more about science than I'll ever grasp.
The discussion seems more about a Newtonian Ego System than about the fundamental questions of the universe.
It’s ironic how the default virtue signal about racism and sexism right now is to attack white men, because to attack someone’s race and gender is the definition of racism and sexism.
“I’m disappointed by the naivete of the questions that you’re communicating,” [Wolfram] grumbles. “I deserve better.”
And maybe that's actually true. Of course, many will find this manner of speech pushes our buttons; coming across as arrogant, megalomaniacal, almost anti-scientific-process.
But that has NOTHING do to with whether this theory is correct or not. When the most obnoxious person you can imagine says 2 + 2 = 4, that doesn't change the fact that 2 + 2 actually does equal 4.
(Of course, if they claim 2 + 2 = 5, their being obnoxious doesn't make that true either.)
Or:
- Schroedinger vs Heisenberg description of QM
- Path integral formulation
- Feynman diagrams (here you can argue that it is not even a formulation, "just a visual representation")
The problem with Stephen Wolfram's theory of everything is not that it does not explain new things. The problem is that (as it is right now) it does not explain new things (he should wish to get there). The problem is that it follows the pattern:
1. Simple graph automata rules.
2. A loose connection with a few physics properties.
3. ???
4. Theory of everything! Profit!!!
I just read a few chapters. And later, I was surprised by the criticism. It turned out that "A New Kind of Science" is not a tongue-in-cheek title (vide "Freakonomics"), but that SW actually believes that it not only is a new kind of science, but his science (Turning, Conway, Feynman, anyone...?).
Some cellular automata simulations are nice. And in general, many concrete in physics are not general, yet are useful e.g. for numerics (with known limitations). They are not considered crackpot (even if not useful), as long as they make clear what they predict, what - not. Any toy model usurping its right to be the TOE, without a sufficient proof (not just the lack of proof it is not) is likely to be crackpotism.
(Side note: for quite a few years I used heavily Mathematica. Only a decade ago I switched to Python, for a few reasons, but mostly for open-source (and quite a few of its implications). Jupyter Notebook, which IMHO was a game-changer for data science in Python, clearly had quite a few inspirations from Mathematica.)
I enjoy toy models, either simplified equations or computational models like CA. I even made a toy model for the spread of corona virus, so I can look at the effects of isolation, death rate, etc. on the spread of infection, and to give my students something to play with during isolation: https://lee-phillips.org/corona/ But I don't claim that this is a serious epidemiological model. Like CAs and the like, I'm hoping that it is fun to play with and might, just possibly, lead to some insight or new idea. But then that new hypothesis would have to be studied using a more serious model.
I suggest them (even if they are not physicists!) looking at the "The Crackpot Index" by John Baez (http://math.ucr.edu/home/baez/crackpot.html) and manually calculating the score for Stephen Wolfram.
Therefore, if materialism is true, then in the beginning must be some simple, but potentially Turing complete substrate, such as a cellular automata or graph thingamabob.
Well, the universe appears to be close to flat, which would indicate an infinite size. And the discreteness of spacetime in QM hasn't been made to work with Relativity, thus the forays into String Theory and Quantum Loop Gravity.