Hacker News Reader

Comment by underdeserver | Hacker News Reader

My problem with Wolfram is that every time I hear him talk or read him, I end up spending 15-20 minutes and come out the other end confused.

I studied physics in my undergrad, I have a graduate degree in math, I've shipped large features on planet-scale systems. I'm no genius, certainly nowhere close to Dr. Wolfram, but I don't think I'm far from your average HN reader in reading comprehension ability.

I've read 35 paragraphs in a post about founding a research institute and I don't know what was said there that couldn't have been conveyed in two or three paragraphs instead.

It is my default opinion that people who use lot word when few word do trick do so to cover up the fact that they have very little content of actual value. That might not be the case with Wolfram, who has quite a track record, but to me the burden of proof is still on him.

View on HN

Wolfram deserves accolades for Mathematica. It's a magnificent achievement and he's kept his company alive and thriving and people use Wolfram (the company) tools everyday to do productive stuff.

BUT... Wolfram (the man) is very much an outsider when it comes to physics. This started 20 years ago when he published a "A New Kind of Science". It was a lengthy but ultimately not useful exploration of cellular automata (some of the fractal-like graphics were pretty cool at the time, but that's about it). Even that was perfectly fine, a rich guy with a vanity hobby doing some weird but interesting stuff, so what?

But lately he's really doubled down on the "theory of everything" stuff with his "hypergraph" idea. To be honest, it IS a bit cringe-worthy. To call this pursuit "The Fundamental Theory of Physics"... geezus, that takes a level of ego that makes one's mouth drop.

But you know what, it's still fine. It's not like he is applying for grants and competing with other worthwhile science for funding. His "staff" for this endeavor are his paid employees. AFAIK, he has no collaborators in academia (and no, multi-hour Twitch streams where he firehoses PHD level math jargon to the general public DON'T COUNT as collaboration).

Now he wants to launch an institute? Still fine by me as long as the funding it comes out of Wolfram's pocket and doesn't detract from more worthwhile non-vanity research.

I suppose you are not familiar with his early career as a physicist, publishing in the field since he was 15 https://en.wikipedia.org/wiki/Stephen_Wolfram#Early_career

That doesn’t tell us much about the quality of the work under discussion.

No. But I'm replying to the OP argument that his physics career began with the book.

No, I was trying to say that the book was when he first demonstrated that he was very much an outsider. Even by then he had been out of academia for a long time, busy building his company and making his fortune.

That was 40+ years ago.

It may or may not be correct, but to get hung up on the ego and the name on the paper is a big mistake. Read the ideas. Maybe they're too weird to understand. But don't condemn them because where they came from. Find the flaws and do it scientifically.

There are only so many hours in a day. If we spend all of them thoroughly debunking known quacks nothing else will ever get done.

Stephen Wolfram hasn’t yet achieved full quack status where it’s best to dismiss him out of hand, but he has been moving in that direction for some time.

It's an interesting idea but it's still very nascent and speculative.

In these last couple of years, Wolfram has been talking about a "universe building kit". He posits that universes (including our own universe) and it's time evolution can be described by relatively simple "hypergraph" that modifies itself and that fully describes physical laws including the concepts that would unify general relativity and quantum mechanics. His thinking is that we just have to somehow "find" that hypergraph and we'll know everything all the way back to "t=0".

Not surprisingly, the physics community doesn't seem interested in dropping whatever they're doing and pursuing this. But like I said, it's OK. The only gripe I would have is if he started trying to getting funding from sources which other scientific researchers use now. The concern is that he would just roll out his obscurantist papers and talks and convince people with the purse-strings to fork over money for this idea that COULD HAVE GONE towards more legitimate pursuits. As it is now, high-energy physics already gets what many consider to be "too much" funding and it's been rather fruitful. This hypergraph stuff... meh.

Try reading his one year update instead[1]

"But now this deflection is not in physical space but in branchial space. The fundamental underlying mathematical structure is the same in both cases. But the interpretation in terms of traditional physics is different. And in what to me is a singularly beautiful result of our models it turns out that what gives the Einstein equations in physical space gives the Feynman path integral in branchial space. Or in other words, quantum mechanics is the same as general relativity, except in branchial space rather than physical space."

Most of this stuff is mind-bending to me the same way statistical quantum mechanics, information theory or higher-dimensional physics are. But I always come away from reading about it with a feeling of "there could be something here, and I'm glad someone is exploring it further"

https://writings.stephenwolfram.com/2021/04/the-wolfram-phys...

It sure sounds promising. Lots of things do. This is why we have peer review and standard publishing practices.

This idea never sat right with me. Only one of Einsteins paper's went through the now standard practice, and he was upset that it did.

Peer review is useful, but let's not pretend it's the only way forward for a theory of everything. It's a very, very modern practice. We did a lot of great physics before it existed.

Nearly everything wolfram does is available to anyone who is interested (including code!).

And just to add on top of this - there is to the best of my knowledge little if any research that shows that the current practice of publication and peer review is actually effective at weeding out bad science. The current replication crisis in psychology and other fields would suggest that it's actually quite bad at it. For a community that is supposed to be based on the ideal of experimentation and falsifiability, scientists are surprisingly willing to just accept at face value, that the current process is sound, without actually scrutinizing this hypothesis.

> The current replication crisis in psychology and other fields would suggest that it's actually quite bad at it.

Not saying that you're necessarily wrong but be careful not to compare apples to oranges: There is a huge difference between fields like math & physics on the one hand and fields like psychology on the other.

Definitely, but both use peer review.

But peer review works completely differently in math & physics than in psychology. In math and (theoretical) physics, practically all results (claims) can in principle be reproduced (tested) by other people. Sure reviewers are usually short on time and won't verify every single detail but at the same time – being experts in their fields – they tend to be extremely good at distinguishing the "easy" arguments from more intricate ones and can usually come up with rough arguments as to why (or why not) a given new result makes sense in light of existing results & literature.

I don't think it is accurate to use problems in the field of psychology to indict peer review in math or physics.

The reviewer was Tolman (or Robertson I think Tolman), and it was arguably one of his most important ones.

It's not about peer review, it's about making predictions. Theories that make predictions that come out true are good ("correct"). Peer review is a poor substitute.

Peer review is just a heuristic people use to spend time more efficiently: You know that if a paper is peer reviewed, it's more likely to be worthwhile spending time actively engaging with it than if it weren't.

But that's only statistical - if you want to quickly screen thousands of research papers and the only signal you have is peer review and maybe citation graph - and doesn't apply to specific papers or the research program of a specific person.

He writes in a very marketing-informed way.

He clearly point to (his) previous work, clearly and repeatedly remembers the reader how amazing is the things he is trying to do, and shameless link (both in text and via hiperlinks) to his own enterprises and current initiatives.

I suppose this is some kind of mental template he uses when writing (think about someone writing a paper: the forma and hidden rules, the necessary and obligatory parts).

Of course I cannot vouch for his science but heck! the man is spending his own money, not trying to sell you anything (other than Mathematica) and has an impressive track record since he was a kid! Let’s give him some space to do his things.

> He writes in a very marketing-informed way.

I'm going to borrow this phrase. I always struggle with dismissing BS without being abrasive.

Thank you for posting this - I read the article too and was left feeling somewhat "dumb" and confused by all the scientific and market speak babble thrown in there. I was doubting my own intelligence for a bit there ... :)

Would you say Wolfram provides low epistemic legibility[] ?

[] https://news.ycombinator.com/item?id=30802349

I really appreciated the concept of "epistemic legibility" as a virtue. Arguments are most valuable when they are clear and concise. It makes them easier to refute if wrong and easier to extend/apply if correct.

As far as I know, the phrase was coined here: https://acesounderglass.com/2022/02/07/epistemic-legibility/

Reply on news.ycombinator.com