So I somehow missed the Stephen Wolfram announcement. I need to get out more (intellectually) while I'm cooped up physically. Has anybody else spent time looking at it? This blog post is hauntingly beautiful: https://writings.stephenwolfram.com/2020/04/finally-we-may-have-a-path-to-the-fundamental-theory-of-physics-and-its-beautiful

I read A New Kind Of Science a while back and found it pretty mind blowing. I saw the announce, but I'm not sure I have anything to contribute at the moment. If I did it would be along the lines of visualization tools.....

I've been working my through it. Yes, it is beautiful. https://twitter.com/seanjtaylor/status/1251238407172272129

Yeah I took a read through this. I found it very compelling. I don’t have enough physics background to know just how special it is for things that look like relativity and quantum mechanics to fall out of the framework, but it seems incredible. I read (most of) A New Kind of Science and got pretty captivated by cellular automata, so I was pretty well-primed to get nerd-sniped by this. It also got me to play around with their notebook environment. The language doesn’t feel good, but it made implementing Langton’s Ant as a graph rewriting rule pretty straightforward.

When you have a hammer, everything looks like a nail. Wolfam has recursive functions, and to him the entire universe looks like a recursive function. Since he is unconcerned with experiments of any kind, he can avoid that intensely distasteful experience of his theories not matching reality. He has a very successful language company and has the luxury of playing around with esoteric theories. We language inventors should all be so lucky. Wolfram is privately owned and makes somewhere between 50 and 100 million a year. I am fairly jealous!

Thanks for posting this, I hadn't see it before! That was an interesting read. It confirms my general view of Stephen Wolfram: he's a genius, an excellent writer, and he constantly oversells his work. What he explains in this essay is that rewrite rules for hypergraphs can produce structures that share the essential properties of space-time as it is traditionally postulated as the starting point of physical theory. It is certainly not obvious that this is possible, and it is even less obvious to demonstrate that this is possible. So for anyone interested in possible computational frameworks for explaining the universe, this is very important work. On the other hand, he clearly says that he has not (yet?) found a set of rewrite rules that is compatible with all the details of our universe as we know it. Which in turn implies that at this stage, he does not have something one could call a scientific model, something that can be used to make testable predictions. There is a clear tradition for such work in the corner of theoretical physics looking at the foundations. String theory is famous for having lots of enthusiastic contributors in the absence of any reasonable prediction after decades of intense research. Time will tell if anything more concrete will come out of this. Wolfram repeatedly points to the biggest obstacle on the way to a computational foundation for physics: computational irreducibility. What this means is, basically, that exploring the consequences of a computational model may require too much time for any practical use, because one would be simulating the universe on a computer necessarily much slower than the universe itself. He seems optimistic that he can escape from that problem, but it isn't very clear to me why.

@Edward de Jong / Beads Project but the universe is a recursive function. I’d rather say a self-modifying recursive function so I’m not sure if I can call that a function since it has sideffects on itself so it isn’t pure but well, something like that - I’d say. I love the idea and support the exploration efforts. Reminds me of my lattice structure and it’s different graphs with different rules in different dimensions, except here the rule is the one that creates while in mine the user creates. I’m not sure if bruteforcing is the right path but I assume it will give them enough data to detect some patterns and rules at evolutionary scale and discern better paths from those patterns.