The Death of Mathematics: How Guild Culture and Artificial Intelligence Are Dissolving the World's Oldest Knowledge Monopoly

So these are Mathematics guys sticking their nose into the universe, with the physics guys.
It could be wrong of course, a perfectly legitimate mathematical argument may not connect to our physical universe but these guys do look at these things.
Penrose and Hawking did this also with their theorems, note spelling.
 
Pinball1970 said:
So these are Mathematics guys sticking their nose into the universe, with the physics guys.
It could be wrong of course, a perfectly legitimate mathematical argument may not connect to our physical universe but these guys do look at these things.
Penrose and Hawking did this also with their theorems, note spelling.
Click to expand...
I am pretty sure that Royal Society article is pure non sense. How can you prove it right or wrong? If not it is non sense.
 
exchemist said:
Then you have no business demanding proof in post 63.
Click to expand...
I don't see the logic here. I stated my instinct (no evidence required by definition). For that specific article, based on Popper's philosophy, you need to have certain way to falsify it. It apparently lacks it therefore it is not serious science, i.e. non sense.
 
Physics & Math Forum.
aispokesperson said:
I am pretty sure that Royal Society article is pure non sense. How can you prove it right or wrong? If not it is non sense.
Click to expand...
Is that based on pure ignorance, or on your not agreeing with some part of the paper?
Do you understand where the paper was leading to?
I would give you my understanding of it, but, "my instinct" ( since that's all you use) is telling me you're not worth the effort of explaining it to.
 
PratherJ said:
Physics & Math Forum.
Is that based on pure ignorance, or on your not agreeing with some part of the paper?
Do you understand where the paper was leading to?
I would give you my understanding of it, but, "my instinct" ( since that's all you use) is telling me you're not worth the effort of explaining it to.
Click to expand...
I am blessed with pure ignorance on this paper's subject. But I can tell it is non sense using common sense.
 
aispokesperson said:
I am pretty sure that Royal Society article is pure non sense. How can you prove it right or wrong? If not it is non sense.
Click to expand...
On what basis do you consider something "non sense" just because, as you imply, it is unfalsifiable? Do you consider everything that is unfalsifiable to be "non sense"? Like other people being conscious? Like the existence of galaxies beyond our observable universe? The love you have for your mother, or wife, or siblings, or some other person.
None of these simple examples are "non sense". Yet they are all unfalsifiable.

So, as Judge Robert Sanders might say: enough of your jibber-jabbering!
 
Anyway, as promised, I have posted a thread on intelligence and machines reporting a problem solved by Paul Edros that has been challenged by Ai.
I say challenged because the mathematics community may find something yet.
I only have the abstract but I also posted a reaction to it from the community which goes into it.
 
Jarek Duda said:
Physics might be also soon solved by AI -
Click to expand...
No, physics is not mathematics, it is empirical. Physics will be "solved" by finding out things about the universe, so in labs, at CERN and using ground based and space telescopes.
In terms of finding trends in the data, scientists already use processor power for that and Ai will feature.
 
While there are lots of unsolved questions in physics, currently mainstream claims there remained to unify GR with SM - both are Lagrangian formalisms, so the question is "just" finding The Final Lagrangian - AI could automatically search for, mainly writing and running simulations to test agreements (mainly with SM+GR), modifying according to disagreements.

Current SM Lagrangian is this gigantic obtained epicycle-style: by guessing and fitting more terms when previous disagree.
We need new Copernican Revolution - to compress its behavior into smaller Lagrangian, also including gravity.
There is focus on Feynman diagrams with particles as abstract points, what is effective perturbative approximation - deeper there are nonperturbative with particles as field configurations, like https://en.wikipedia.org/wiki/Skyrmion - with simple looking nonperturbative Lagrangians, but of extremely complicated consequences - AI could automatically explore to find the right one.

1782447377348.png
 
Last edited:
Jarek Duda said:
While there are lots of unsolved questions in physics, currently mainstream claims there remained to unify GR with SM - both are Lagrangian formalisms, so the question is "just" finding The Final Lagrangian - AI could automatically search for, mainly writing and running simulations to test agreements (mainly with SM+GR), modifying according to disagreements.
Click to expand...
It is probably time to look for a Hamiltonian for the "Theory of Everything," rather than a Lagrangian.;)
Regarding AI, two things can be said: first, what we might wish for (conservatism); and second, what is already -de facto- not the future but the present (affecting not just mathematics and physics, but absolutely everything) and is something we can no longer change. Progress used to be a linear function; now it is an $$x^x$$ function -it multiplies itself.
 
Last edited:
Hamiltonian is indeed often more convenient e.g. to optimize field configuration of particle as topological soliton, but more sophisticated scenarios need Lagrangian for Euler-Lagrange/least action/Feynman ensembles.
Also Lagrangian is usually given as it allows to derive Hamiltonian with Legendre transform, the opposite is more difficult.

Regarding AI to search for deeper physics, the question is basically: "is this small nonperturbative Lagrangian effectively described close to this huge of Standard Model + gravity"?
AI is approaching capability to investigate such difficult question - e.g. in https://github.com/openwave-labs/openwave/blob/main/MODELS.md already testing in simulations Lagrangians provided by humans ... if insufficient, AI might search for modifications, new Lagrangians, e.g. in evolutionary way.

So I suspect new Copernican revolution compressing this gigantic epicycle-style Lagrangian, into a small one with few parameters allowing to derive all ~30, is very close now.
E.g. PMNS neutrino oscillation parameters - previously available only from these gigantic experiments, and recently AI claims to derive them from such deeper Lagrangian: https://github.com/openwave-labs/op...stal/research/findings/m5_11_pmns_findings.md

1783475878709.png
 
Last edited:
Jarek Duda said:
Regarding AI to search for deeper physics, the question is basically: "is this small nonperturbative Lagrangian effectively described close to this huge of Standard Model + gravity"?
AI is approaching capability to investigate such difficult question - e.g. in https://github.com/openwave-labs/openwave/blob/main/MODELS.md already testing in simulations Lagrangians provided by humans ... if insufficient, AI might search for modifications, new Lagrangians, e.g. in evolutionary way.
Click to expand...

It might turn out to be like the formula for solving a fourth-degree equation -several pages of fine print.
 
You must log in or register to reply here.
Back
Top