Keep in mind that Wolfram has a tendency to over exaggerate his claims and their import. He is undoubtedly a great physicist, but everything I’ve seen indicates this is more or less a...
Keep in mind that Wolfram has a tendency to over exaggerate his claims and their import. He is undoubtedly a great physicist, but everything I’ve seen indicates this is more or less a reformulation of known results from computational geometry interspersed with hand waving arguments about deeper importance.
For background on Wolfram’s legendary hubris, I’d recommend this American Mathematical Society review of Wolfram’s A New Kind of Science (2002). My impression is that this current work is really...
For background on Wolfram’s legendary hubris, I’d recommend this American Mathematical Society review of Wolfram’s A New Kind of Science (2002). My impression is that this current work is really just an extension of the ideas presented in that book. Nevertheless, it’s notable that he has other physicists working with him, so it’ll be interesting to see how his ideas play about.
I'm enjoying these more than I dare to admit. It seems obvious but it's just wonderful to see it spelled out like this, I feel like I better understand "cranks" now. Jordan Peterson also perfectly...
I'm enjoying these more than I dare to admit. It seems obvious but it's just wonderful to see it spelled out like this, I feel like I better understand "cranks" now. Jordan Peterson also perfectly matches this pattern for psychology.
"Wolfram has a tendency to over exaggerate his claims" compares favorably with British Airways Flight 9's infamous "we have a small problem; all four engines have stopped" for dramatic...
"Wolfram has a tendency to over exaggerate his claims" compares favorably with British Airways Flight 9's infamous "we have a small problem; all four engines have stopped" for dramatic understatement. I would strongly recommend any lay person ignore everything he says. If any of this stuff gets published in peer-reviewed journals and picked up by the broader math and physics communities, then it's worth taking note of.
It's a moderately delicate line to tread between alerting laypeople that someone's claims are embellished and telling them to completely ignore them. There is definitely a trend in popular science...
It's a moderately delicate line to tread between alerting laypeople that someone's claims are embellished and telling them to completely ignore them. There is definitely a trend in popular science literature these days that portrays the physics community as stuck on old ideas and hostile to new ones, despite "no progress" in X decades, where X depends on what the author views as relevant physics. You see people like Sabine Hossenfelder promoting this viewpoint, which trickles into the passive observer's mental picture of what the field is these days. If you then tell that observer to ignore "new ideas" because they haven't been reviewed by the "community", that gets filtered through their skepticism of the physics community and the result is often more attention paid to the fringe ideas. Obviously this is all anecdotal, but I find it's usually better to measured in criticism, especially directed towards someone like Wolfram who is clearly not a crackpot.
As for the overall trend described above, I'm not sure what the answer is. It is easy to say that particle physics since the 70s has not produced any new or unexpected results (no closer to getting testable predictions from string theory, no supersymmetry at the LHC, etc), but this isn't really the case. Unfortunately it's much easier for people who are potentially not acting in good faith (cough Hossenfelder) to lambast things as misguided than it is for experts in those subfields to explain the nuance.
Anyway this is all a bit off-topic, but suffice to say I am not that excited about these new Wolfram developments, both in that I am not sure they are new or that they are even developments.
I just finished reading the whole thing and I'm intrigued, he's managed to explain everything from relativity and quantum mechanics to dark matter and black holes, along with what gravity actually...
I just finished reading the whole thing and I'm intrigued, he's managed to explain everything from relativity and quantum mechanics to dark matter and black holes, along with what gravity actually is, all from nothing more than the concept of a 'relationship'. The problem is, this is only being done in toy models. There's a staggering amount of work to do reconciling this with all of known physics and zeroing in on the right model that accurately represents our universe - if that's even possible. I'll give him this - it's a truly novel way of tackling the problem. This isn't a theory of everything so much as it is a different toolset to get there that shows intriguing results.
Considering that the standard approach to advancing physics has become a fossil which has saddled us with decades of no real progress, I'm in the camp that says any new approaches, tools, and toys are a good thing.
We need to get Steven in a room with Eric Weinstein. Eric's theory of everything (which he finally shared last week in his Portal podcast) takes the same kind of approach Steven is using. Step back from all of the theories and start with a simple, basic element, then iterate that until you can recover something that looks like all known physical systems from it - and pray that on the way, it puts some new ideas in your head.
It certainly shows the power of Wolfram's computational language for visualization, but I want to push back a bit at the claim that there's a ton here that's new and novel. This isn't to say it's...
I just finished reading the whole thing and I'm intrigued, he's managed to explain everything from relativity and quantum mechanics to dark matter and black holes, along with what gravity actually is, all from nothing more than the concept of a 'relationship'. The problem is, this is only being done in toy models. There's a staggering amount of work to do reconciling this with all of known physics and zeroing in on the right model that accurately represents our universe - if that's even possible. I'll give him this - it's a truly novel way of tackling the problem. This isn't a theory of everything so much as it is a different toolset to get there that shows intriguing results.
It certainly shows the power of Wolfram's computational language for visualization, but I want to push back a bit at the claim that there's a ton here that's new and novel. This isn't to say it's not interesting, but as I mentioned in a comment above Wolfram tends to be very loose in distinguishing what is a new result and what is known (he claims to have "discovered" a lot of things, very few of which he legitimately was the first one to discover). For example, much of what he details in this announcement is known from loop quantum gravity, causal sets, and the Ads/CFT correspondence. His exposition on deriving geometric results from hypergraphs use pretty much standard definitions from computational geometry, and similarly his definitions of mass and energy are pretty much the definitions from general relativity as codified in a stress-energy tensor. It's not a big surprise then, given these definitions, that he gets special relativity back out. It's sort of baked in. This is kind of the case with a lot of what is going on here - he defines things in a standard way, and then seems surprised when standard things pop out. For example, he says that curvature in his theory can be defined as variation in local dimensions which is fine, but also obvious given that he defined effective dimension as the ration of local volume to radius. His results regarding QM are a bit more of the same. The connection between quantum complexity and geometry is the subject of the aforementioned Ads/QFT correspondence which has been one of the most fruitful aspects of high energy gravitational theory in the last 15ish years.
Again, none of this is to say that this isn't interesting. But Wolfram is incredibly frustrating in his personal embellishments, lack of distinction between what is new and what is a standard result, and quite frankly his lack of acknowledgement of current literature and work. This can be even more frustrating when the view that
Considering that the standard approach to advancing physics has become a fossil which has saddled us with decades of no real progress, I'm in the camp that says any new approaches, tools, and toys are a good thing.
is as prevalent as it is. There are obviously real critiques with the state of high energy theory of the bast few decades, but it is far, far from no "real progress". This becomes a problem when people like Wolfram don't acknowledge drawing on this progress for their own work and developments, and ultimately furthers obscures advances in fundamental physics.
After reading it, I’m trying to reconcile if this is really novel or not. As far as I can tell, Wolfram is essentially offering a framework for modeling. While it’s certainly neat, what I don’t...
After reading it, I’m trying to reconcile if this is really novel or not. As far as I can tell, Wolfram is essentially offering a framework for modeling. While it’s certainly neat, what I don’t see is the evidence that these models are predictive. That is, the toy models can be used to explain physics that we ready have different models for, and it’s possible Wolfram’s models are not elegant. But, what I’m not seeing is the predictive power. Having a modeling framework is different than having a useful model.
This is something I run into in my field of natural language processing. I can build lots of models within a given framework and some parameter space. But if I have no objective way to discern which model instance is preferable to another, the models don’t do me much good. Wolfram starts to dig at these issues of model utility and interpretability with the section discussing dimensionality. But, he raises more questions than he answers by explaining how to create models of space that have non-integer dimensions (and doesn’t even explain what these models might represent).
Graphs are very powerful things in the abstract. It’s not surprising that they are very expressive, and it’s also not surprising that we can think about the universe as some kind of automaton. Graphical models are also very convenient for computational reasons, too. But, I think there is easily a danger of ascribing too much hype to the notion that just because hypergraphs are useful, that they are surely the be all and end all to theoretical physics.
This is very cool. I'm only halfway through the article (it's pretty dense in a good way), but I know I'm going to come back and finish it after I get some sleep. Thanks for sharing this. One of...
This is very cool. I'm only halfway through the article (it's pretty dense in a good way), but I know I'm going to come back and finish it after I get some sleep. Thanks for sharing this. One of the largest mindfucks I've had in my life is in thinking about theories and philosophies of space and time, and while this continues to boggle, it makes a little bit more sense so far than I'm used to.
The idea of a tiny model from scratch having both relativity and quantum mechanics fall out of it as emergent properties is really exciting to me. I've long thought that we need models like this...
The idea of a tiny model from scratch having both relativity and quantum mechanics fall out of it as emergent properties is really exciting to me. I've long thought that we need models like this so we can build up some intuition around these theories, even if the model turns out not to be fully true to reality. I'm surprised this sort of thing hasn't been done more.
Also whether or not this ends up being useful as a model of reality, it's a super interesting alternative to cellular automata. It avoids the issues of cellular automata presupposing a rigid grid system and global synchronous timestep. I'm interested in making some simulations based on this and seeing if I can produce anything interesting on a small scale. I've read the whole post and I've been diving into some of the linked pages. I've always been fascinated by gliders and other structures in Conway's Game of Life.
Unfortunately as a non-physicist, all I can tell is that Wolfram is excited about it. Hopefully someone else who knows something will tell us if it turns out to be useful.
Unfortunately as a non-physicist, all I can tell is that Wolfram is excited about it. Hopefully someone else who knows something will tell us if it turns out to be useful.
So is the ELI5 version that the universe is procedurally generated from a seed, for lack of a better term, and we know what some of the rules are for generation, but not the seed nor all the...
So is the ELI5 version that the universe is procedurally generated from a seed, for lack of a better term, and we know what some of the rules are for generation, but not the seed nor all the rules?
cc: @hungariantoast, I think it reminds you of procedural generation because that's what he seems to be arguing for. It makes sense for the universe at some level, but the argument that the laws of physics themselves were generated in such a fashion is a lot to take in.
Keep in mind that Wolfram has a tendency to over exaggerate his claims and their import. He is undoubtedly a great physicist, but everything I’ve seen indicates this is more or less a reformulation of known results from computational geometry interspersed with hand waving arguments about deeper importance.
For background on Wolfram’s legendary hubris, I’d recommend this American Mathematical Society review of Wolfram’s A New Kind of Science (2002). My impression is that this current work is really just an extension of the ideas presented in that book. Nevertheless, it’s notable that he has other physicists working with him, so it’ll be interesting to see how his ideas play about.
That was a delicious read, lol. Starting from dry skepticism and suddenly comparing him to L. Ron Hubbard on page 5.
They get worse yet. Here is a much harsher rebuke.
I'm enjoying these more than I dare to admit. It seems obvious but it's just wonderful to see it spelled out like this, I feel like I better understand "cranks" now. Jordan Peterson also perfectly matches this pattern for psychology.
"Wolfram has a tendency to over exaggerate his claims" compares favorably with British Airways Flight 9's infamous "we have a small problem; all four engines have stopped" for dramatic understatement. I would strongly recommend any lay person ignore everything he says. If any of this stuff gets published in peer-reviewed journals and picked up by the broader math and physics communities, then it's worth taking note of.
It's a moderately delicate line to tread between alerting laypeople that someone's claims are embellished and telling them to completely ignore them. There is definitely a trend in popular science literature these days that portrays the physics community as stuck on old ideas and hostile to new ones, despite "no progress" in X decades, where X depends on what the author views as relevant physics. You see people like Sabine Hossenfelder promoting this viewpoint, which trickles into the passive observer's mental picture of what the field is these days. If you then tell that observer to ignore "new ideas" because they haven't been reviewed by the "community", that gets filtered through their skepticism of the physics community and the result is often more attention paid to the fringe ideas. Obviously this is all anecdotal, but I find it's usually better to measured in criticism, especially directed towards someone like Wolfram who is clearly not a crackpot.
As for the overall trend described above, I'm not sure what the answer is. It is easy to say that particle physics since the 70s has not produced any new or unexpected results (no closer to getting testable predictions from string theory, no supersymmetry at the LHC, etc), but this isn't really the case. Unfortunately it's much easier for people who are potentially not acting in good faith (cough Hossenfelder) to lambast things as misguided than it is for experts in those subfields to explain the nuance.
Anyway this is all a bit off-topic, but suffice to say I am not that excited about these new Wolfram developments, both in that I am not sure they are new or that they are even developments.
I just finished reading the whole thing and I'm intrigued, he's managed to explain everything from relativity and quantum mechanics to dark matter and black holes, along with what gravity actually is, all from nothing more than the concept of a 'relationship'. The problem is, this is only being done in toy models. There's a staggering amount of work to do reconciling this with all of known physics and zeroing in on the right model that accurately represents our universe - if that's even possible. I'll give him this - it's a truly novel way of tackling the problem. This isn't a theory of everything so much as it is a different toolset to get there that shows intriguing results.
Considering that the standard approach to advancing physics has become a fossil which has saddled us with decades of no real progress, I'm in the camp that says any new approaches, tools, and toys are a good thing.
We need to get Steven in a room with Eric Weinstein. Eric's theory of everything (which he finally shared last week in his Portal podcast) takes the same kind of approach Steven is using. Step back from all of the theories and start with a simple, basic element, then iterate that until you can recover something that looks like all known physical systems from it - and pray that on the way, it puts some new ideas in your head.
It certainly shows the power of Wolfram's computational language for visualization, but I want to push back a bit at the claim that there's a ton here that's new and novel. This isn't to say it's not interesting, but as I mentioned in a comment above Wolfram tends to be very loose in distinguishing what is a new result and what is known (he claims to have "discovered" a lot of things, very few of which he legitimately was the first one to discover). For example, much of what he details in this announcement is known from loop quantum gravity, causal sets, and the Ads/CFT correspondence. His exposition on deriving geometric results from hypergraphs use pretty much standard definitions from computational geometry, and similarly his definitions of mass and energy are pretty much the definitions from general relativity as codified in a stress-energy tensor. It's not a big surprise then, given these definitions, that he gets special relativity back out. It's sort of baked in. This is kind of the case with a lot of what is going on here - he defines things in a standard way, and then seems surprised when standard things pop out. For example, he says that curvature in his theory can be defined as variation in local dimensions which is fine, but also obvious given that he defined effective dimension as the ration of local volume to radius. His results regarding QM are a bit more of the same. The connection between quantum complexity and geometry is the subject of the aforementioned Ads/QFT correspondence which has been one of the most fruitful aspects of high energy gravitational theory in the last 15ish years.
Again, none of this is to say that this isn't interesting. But Wolfram is incredibly frustrating in his personal embellishments, lack of distinction between what is new and what is a standard result, and quite frankly his lack of acknowledgement of current literature and work. This can be even more frustrating when the view that
is as prevalent as it is. There are obviously real critiques with the state of high energy theory of the bast few decades, but it is far, far from no "real progress". This becomes a problem when people like Wolfram don't acknowledge drawing on this progress for their own work and developments, and ultimately furthers obscures advances in fundamental physics.
After reading it, I’m trying to reconcile if this is really novel or not. As far as I can tell, Wolfram is essentially offering a framework for modeling. While it’s certainly neat, what I don’t see is the evidence that these models are predictive. That is, the toy models can be used to explain physics that we ready have different models for, and it’s possible Wolfram’s models are not elegant. But, what I’m not seeing is the predictive power. Having a modeling framework is different than having a useful model.
This is something I run into in my field of natural language processing. I can build lots of models within a given framework and some parameter space. But if I have no objective way to discern which model instance is preferable to another, the models don’t do me much good. Wolfram starts to dig at these issues of model utility and interpretability with the section discussing dimensionality. But, he raises more questions than he answers by explaining how to create models of space that have non-integer dimensions (and doesn’t even explain what these models might represent).
Graphs are very powerful things in the abstract. It’s not surprising that they are very expressive, and it’s also not surprising that we can think about the universe as some kind of automaton. Graphical models are also very convenient for computational reasons, too. But, I think there is easily a danger of ascribing too much hype to the notion that just because hypergraphs are useful, that they are surely the be all and end all to theoretical physics.
This is very cool. I'm only halfway through the article (it's pretty dense in a good way), but I know I'm going to come back and finish it after I get some sleep. Thanks for sharing this. One of the largest mindfucks I've had in my life is in thinking about theories and philosophies of space and time, and while this continues to boggle, it makes a little bit more sense so far than I'm used to.
The idea of a tiny model from scratch having both relativity and quantum mechanics fall out of it as emergent properties is really exciting to me. I've long thought that we need models like this so we can build up some intuition around these theories, even if the model turns out not to be fully true to reality. I'm surprised this sort of thing hasn't been done more.
Also whether or not this ends up being useful as a model of reality, it's a super interesting alternative to cellular automata. It avoids the issues of cellular automata presupposing a rigid grid system and global synchronous timestep. I'm interested in making some simulations based on this and seeing if I can produce anything interesting on a small scale. I've read the whole post and I've been diving into some of the linked pages. I've always been fascinated by gliders and other structures in Conway's Game of Life.
Unfortunately as a non-physicist, all I can tell is that Wolfram is excited about it. Hopefully someone else who knows something will tell us if it turns out to be useful.
So is the ELI5 version that the universe is procedurally generated from a seed, for lack of a better term, and we know what some of the rules are for generation, but not the seed nor all the rules?
cc: @hungariantoast, I think it reminds you of procedural generation because that's what he seems to be arguing for. It makes sense for the universe at some level, but the argument that the laws of physics themselves were generated in such a fashion is a lot to take in.