This Theory of Everything Could Actually Work: Wolfram’s Hypergraphs

Mathematician and Computer Scientist Stephen Wolfram wants to do no less than revolutionising physics. He wants to do it with computer code that gives rise to all the fundamental laws of nature that we know and like -- and maybe more. Unfortunately, Einstein’s theories of general relativity inherently clash with how computers work. And yet, he and his team might have found a clever way around this problem.

🔗 Join this channel to get access to perks ➜

#science #physics

[mastergshotgun]

"It could actually work" Isn't this the nicest thing ever Sabine ever said about a new theory?!? Damn! The dude must be stellar!

[jimlaheta]

It's refreshing to see that Wolfram actually listened to the flaw in his theory and found someone to help address them. Good to see that Gorard used the work from the causal sets people instead of reinventing it. Even if the theory/direction is shown to just not work, at least we would know what to not work on, which would already make it better than string "theory".

[WerdnaGninwod]

Sabine didn't mention something that I thought was most curious about Wolfram and Gorard's work, which was in the way they worked with hypergraph rewriting methods.
Firstly, by choosing hypergraphs, they were allowing all possible topologies, and secondly, rather than choosing some specific set of rewriting rules that they thought might work, they chose instead, to integrate across all possible hypergraph rewriting rules.
So, they're not imposing any topological structure and they're not imposing any structure to change over time, and yet structure emerges nevertheless, which is astounding.
Many rewrite rules are computationally equivalent - so they reduce to their simplest form for predictive purposes.
Many rewrite rules produce no structure at all, and so they can be ignored as background noise.
Many rewrite rules produce only momentary structure that self destructs - think like virtual particle pairs that emerge and self-cancel.
Many rewrite rules produce structure that is computationally reducible - and hence the kind of structure we focus on in macroscopic physics where we can predict the outcome, because it can be computed faster than the system that actually enacts it.
Many rewrite rules produce structure that is computationally irreducible - like we see in quantum physics, where there is a probabilistic distribution of potential outcomes the could be determined through Feynman's sum over path integral approach, but which could never be computed faster than the system itself operates.
Essentially though, the structure we observe, is the recurring patterns that emerge from the rewriting rules that actually do produce recurring structure, while everything else naturally falls away.

[charlievane]

I think Wolfram's line of reasoning makes sense, in the sense that he's looking for a pattern from which the laws of physics are an emergent behaviour

[Patrick-rnwu]

This is high praise coming from Sabine. Didn't even get interrupted by Elon on the phone.

[gaborbencsik]

Thank you, for this video, and especially for giving Jonathan Gorard credit for his part in this work, he certainly deserves the attention.

[Helmutandmoshe]

"Wolfram is a curious case because he's not your average crank." - in his teenage years at Caltech, he was publishing research papers in theoretical elementary particle physics that were world class. He was prepped to become the next great theorist and then he changed directions. He's not really a mathematician, he's always been a theoretical physicist at heart and his mathematical approach was a way to get at the patterns deep in nature. It has been a many decades long project and even though I don't think it will be able to derive physics from these supposedly deeper principles, it is an astoundingly original and interesting effort - from a genius.

[xaviercheyron]

As i understand it, and imho this is the main reason physicists should have a look at his work, part of Wolfram's genius in his approach is to remove the observer from the analysis and to study the hypergraph as a whole. This allows to sidestep our observer biais when considering motion and other stuff (Since as observers we are part of the graph). His thesis is that the laws of physics are emerging properties of the hypergraph AND that we only can experience them as observers from within the graph. So we can experience the emerging effects, but can't experience the underlying principles by nature.

That's a level of abstraction i initialy didn't think we would need but it makes sense given how complex systems behave.

[barthbrothers]

The realisation that “Nobody knows what they are talking about…” is a very good starting point👍 A clever physicist always - well, most of the time - becomes more insightful when realising that we are just looking through the keyhole of a door leading into an even bigger room with even more doors….

[TheEliaSenn]

I did stats classes in college and there were two professors I really enjoyed. They had different approaches to fitting a model. One was very fond of running monte carlo simulations and working out a model from the observed data. The other would like to work out the math for the model they theorized it would be, then tested it afterwards. A machine gun vs a sniper rifle. Either approach would probably work, and it was more appropriate to use one method over the other with certain problems. I feel like Wolfram is simply taking a different approach from a 'traditional' physicist because the bulk of his life's work wasn't in traditional physics. Perhaps physicists had a hammer in their hand and saw the problem as a nail, and Wolfram has a screwdriver in his hand and sees the problem as a screw. Time will tell which approach bears more fruit.

[rich]

"The story so far:
In the beginning the Universe was created.
This has made a lot of people very angry and been widely regarded as a bad move." - Douglas Adams

[puzzzl]

I've been fascinated with Wolfram since the release of his much-lambasted A New Kind of Science decades ago, and it's only your videos and discussions of the academic world that have helped me understand his situation. Wolfram is famous for his ego and his radical pronouncements ("A new kind of science!!!1"), but I realize now, what choice does he even have to be heard in a world full of people with their fingers in their ears? His work challenges the academic consensus, and so everyone just laughs at him. I've got the utmost respect for him continuing this work for so long now despite the hostilities. The world needs more original thought.

[kevinvanhorn]

One of the mistakes I see critics of Wolfram's approach make is assuming that the computation steps correspond to increments of time. They don't, and this was made clear even in Wolfram's 2002 book (A New Kind of Science). The other mistake is failing to understand that observations from "outside" the hypergraph are not possible; the observer is itself a substructure of the hypergraph, and this affects what can and cannot be observed.

[jkoblivion]

I'm obsessed with this Channel. Not only it's content but also the presence of Sabine Hossenfelder. She gives so much structure to physics. Thank you from Portugal.

[GeoffPlitt]

Sabine, I've admired your skepticism of junk science so much, and so I was pleased to see you're also encouraged by Wolfram. I think he's onto something, but you're right his work is so hard to understand. This was a great video. Please never stop doing them!

[DrummerRF]

I also love Jonathans attitude towards it. Even if the computational approach doesnt quite work it may still produce mathematics that does reveal new ideas, for example the fact that curvature can also be a measure of non-integer Hausdorf dimensionality. Stuff like that. The way they are incorporating quantum mechanical in their models might actually help teach us a lot! More to them

[stephanematis]

As a Mathematica user since the early 90s, nothing really ever surprises me about Stephen Wolfram's ideas and pursuits.

[chipkrug]

Great explanation of how computational, incremental theories like Wolfram's are, like quantum mechanics, incompatible with general relativity. Reminds me of a notion I had: that Xeno's paradox can be explained by the idea that nature is discontinuous, made up of discrete elements of space and time.

[karl]

You misunderstand at 2:45. His theory doesn’t have time only a sequence of events. Time is an emergent phenomena that is used to make approximate models, but likely makes it harder to understand the fundamental rules.

[ohprudence]

I worked as a freelance technical writer for Mathematica in 1990 and have a lot of respect for Wolfram.