- Simple computational rules can produce incredibly complex behavior.
- Wolfram's hypergraph model derives Einstein's equations from network rewriting.
- Dark matter may be microscopic random motion in space itself, not actual matter.
Stephen Wolfram has a confession: despite building the computational tools that power much of modern science, artificial intelligence hasn't impressed him.
The physicist and founder of Wolfram Research has been hunting for a theory of everything since his days at Caltech, and he believes the answer lies not in neural networks but in something far more fundamental.
In this video from New Scientist, Wolfram explains his breakthrough approach to understanding what he calls "the machine code of the universe."
Key figure
3.01 / 2.99
Predicted dimension fluctuations in space – not exactly three – that could confirm Wolfram's physics model
The Universe Runs on Simple Rules
Wolfram's insight centers on "rulology" - the study of simple computational rules and their complex outcomes. Take a grid of black and white cells with basic rules about how each cell changes color based on its neighbors. Run these rules millions of times, and something unexpected happens.
"I was wrong," Wolfram admits about his initial assumptions. "Even very simple rules produce incredibly complicated behavior."
This discovery led him to a radical proposition: space itself might be made of discrete elements connected by relationships, like nodes in a vast network. Everything we perceive - particles, electrons, quarks - could be features of this underlying spatial structure.
From Hypergraphs to Einstein's Equations
Wolfram represents this network as a "hypergraph" where multiple nodes connect through relationship blobs rather than simple lines. The key insight came in 2020 when he realized time emerges from the computational rewriting of these spatial networks.
What is a hypergraph?
A hypergraph is a network where a single connection (called a hyperedge) can link three or more nodes at once, unlike ordinary graphs where edges connect just two points. Wolfram uses hypergraphs to model space itself: the nodes are the building blocks of space, and the hyperedges are the relationships between them. As these connections are rewritten by simple rules, the familiar geometry of the universe – including time – emerges.
In a hypergraph, generalized edges may connect to three or more nodes. Source: Wikipedia. See also Wolfram's page on graph types.
"It made really big breakthrough in understanding the machine code of the universe," he explains. When you scale up from microscopic network rewriting to large-scale behavior, you get Einstein's equations - the same way molecular motion produces fluid mechanics.
The implications are staggering. Wolfram believes this approach explains why general relativity and quantum mechanics work the way they do, something he "didn't think was possible."
Dark Matter as Spatial Heat
One of Wolfram's most intriguing predictions involves dark matter. He suspects it's not matter at all, but rather the "space-time analog of heat" - microscopic random motion in the structure of space itself.
This echoes a historical mistake in 19th-century physics, when scientists invented "caloric fluid" to explain heat before realizing it was just molecular motion.
There are things which people almost did 100 years ago and got stuck. Now we can unstick them. That's really magic.
Stephen Wolfram
Testing these ideas requires detecting dimension fluctuations in space - regions that are 3.01 or 2.99 dimensional rather than exactly three. Wolfram's team is working on experimental approaches to find these signatures.
Why AI Falls Short
Despite the current AI boom, Wolfram remains unimpressed by large language models for fundamental physics.
"No AI has really impressed me," he states, explaining that neural networks excel at human-like tasks but fail when confronted with computational irreducibility - problems requiring irreducible amounts of computation to solve.
LLMs can search literature and find patterns humans miss, but they can't shortcut the universe's own computational work.
LLMs can search literature and find patterns humans miss, but they can't shortcut the universe's own computational work. The real value lies in using AI to humanize computational discoveries, helping identify which mathematical theorems humans might actually care about.
Wolfram continues his quest not from academic pressure but from pure fascination.
As he puts it, when fundamental insights start fitting together, "you can't not do that."
Fact Check: Claim-by-Claim Verification Verified
The recap accurately represents Stephen Wolfram's statements in the New Scientist video and aligns with his published writings on the Wolfram Physics Project.
Commentary
- The recap matches available transcript and Wolfram's consistent project descriptions.
- Wolfram's ideas remain unproven speculation, accurately presented as such.
Sources used for verification
Academic/Peer-reviewed:
- Finally We May Have a Path to the Fundamental Theory of Physics - writings.stephenwolfram.com
- The Ruliology of Lambdas - writings.stephenwolfram.com
- Wolfram Physics Project - wolframphysics.org
Other reliable sources:
- No AI has impressed me (New Scientist video) - youtube.com
Fact-checked by Perplexity Sonar Pro on 2025-12-27
