Welcome to 2020, where new theories of everything are announced via livestreamed notebook tutorials

Ok going to live-tweet my impressions as I watch

Wolfram defines everything as emerging from a (hyper)graph that evolves via cellular automata-like local update rules

Defines properties of the emergent geometery using standard concepts from discrete differential geometry

Time is considered as the updates to the geometry, particles are irregularities in the geometry that are preserved through updates and can propagate via the basic rules.

Sounds similar to qasiparticles in condensed matter

Proceeds to defining a partial order on different possible updates to the hypergraphs, can create a causal ordering between "events" (graph updates)

Proceeds to some weird sleight of hand "deriving" special relativity by literally applying a boost to a casual graph, waving hands saying "observers experience a different foliation", and then saying "look, it's like a Lorenz boost!". Seemed like s circular argument to me

Claims energy= flux of edges through hyperplane. That one can derive E=mc^2 from this... Ok

Says casual invariance necessarily implies special relativity

"time is just the computational progression of applying update rules. We as observers perceive the causal graph, its causal invariance leads to the independence of reference frames and hence special relativity"

Derives energy and momentum as fluxes through spacelike and timelike hyperplanes... So literally the definition of the stress-energy tensor? Bruh.

Einstein equation follows from causal invariance and from the properties of the pseudorandom processes behind the generation of these graphs.

"Let's talk about the universe before we get to quantum mechanics."
Me:

Any pinches in the spatial or causal geometry ---> wormholes.

Ok.

"we could define curvature as a variation in local effective dimension"

Me: No shit bruv you defined effective dimension as ratio of local volume to radius

Ok now we're temporarily back to vanilla cellular automata - identifies an analogue of what is a particle in the hypergraphs. Admits he has no idea how particles work in his models but he hopes he "can simulate then with computers".

Man, this is turning into a really long ad for Mathematica.

there are some parts of the computation that become reducible - apparently how one derives GR and QM from these hypergraph automata. Proceeds to wildly speculate about this approach being able to understand dark matter or predict masses of particles

Ohh snap time for quantum mechanics - here we go

Name drops "my friend Dick Feynman". Good start.

The paths in a Feynman path integral are analogous to the paths through hypergraph update space.

Defines "quantum oberservers" and "quantum reference frames". Measurement yields a contraction of the foliations. The spread of the consistent history is "like the spread of decoherence". Pinching of foliations "analogous to coordinate singularity" as in BH's

"quantum computing folks are trying to "freeze time" to keep a pure state fixed. Effectively like creating a black hole."

Defines "branchial space", can define a partial ordering over possible updates to hypergraphs (multiway graph). Hypergraphs "correspond to quantum states". Connections = "a map of quantum entanglements"

So geometry is entanglement, hmm, where have I heard that before

Mentions geometry of quantum states and geodesics as paths between quantum states... duality between quantum complexity and geometry, again, totally original here

(aside: for those not familiar, entanglement as geometry is a standard concept in AdS/CFT and the duality between quantum complexity and geometry has been thoroughly explored there)

Geodesics between quantum states are affected by the presence of energy (also defined as flux of causal edges for his version of QM)

"vectors in branchial space determine the phase of a path in the Feynman path integral"

"The analogue of the Einstein equations is the path integral"

"the uncertainty principle comes from geometry, non-commuting operators are analogous to non-commuting covariant derivatives"

"you can make entanglement cones in branchial space, these are analogous to light cones"
Speed of light=maximum entanglement speed

My thoughts: no ; you can't entangle d.o.f.s that commute, Feynman propagator will bound speed of entanglement, itself bound by speed of light

2 hours in : finally mentions AdS/CFT and ER=EPR. So he admits he's familiar with that literature?! He seems to have borrowed TONS of concepts from that scientific community. Took two hours to even acknowledge its existence.

Ok, one interesting point here: vacuum energy is not a problem in this theory as virtual particles *make* space, just unobserved fluctuations in the hypergraph geometry I guess

Wolfram: elementary length is 10^-93 meters

wayyyy tinier than Planck length (10^-35)

I guess there's never ever going to be a way to even test his theories at that energy scale so he can say wtv he wants ¯\\_(ツ)_/¯

K I think we're done with the high-level tour. TLDR: everything emerges from hypergraphs mutate (cellular automata-like), keeping track of mutations gives causal order, latter yields SR. QM consistent histories forces syncing of foliations (causal order). Geometry = entanglement?