In a numerical coincidence, some see evidence for string theory

(I am a person who knows some physics but not a lot).

If any of this stuff turned out to be true, what would the lab experiments and positive outcomes look like?

I'm accustomed to particle accelerators and all sorts of other physical experiments, but I really struggle to relate these sorts of theories to anything that could be done in the lab, and if the theories were predictive, what sort of applied science could be done? No wide-eyed "we could make space elevators and travel FTL", just realistic extrapolation.

There's a good reason why most people who specialize in quantum gravity work in string theory: violating any axioms of string theory leads to theories which are self-evidently inconsistent with themselves and/or experiment. It's worth spending time trying to understand why, but unfortunately a lot of air is sucked out of the room by Woit's petulant negativity. It makes for fun podcasts but I'm glad researchers are willing to actually sit and think about the consequences of physics that doesn't match our intution, even when it's hard.

And indeed this quote misses the point. The point is that there aren't "additional" effective theories of gravity that are self-consistent but couldn't be explained by string theory. This isn't a smoking gun that string theory is correct, but if the opposite were true, it would have suggested that string theory was missing something.

So it's incorrect to conclude from this evidence that string theory is "consistent with everything"; it's merely consistent with the range of things that are consistent, and no more and no less. (Suspicious! But not conclusive.) One can complain that this isn't predictive, and indeed if you want to look for predictive power you look at other physical quantities. String theory makes very precise and unique predictions for the way that e.g. gravitons scatter off each other. The fact that humans don't (yet) have the ability to do such experiments is regrettable from a fundamental physics perspective, but it's a problem with humans, not string theory. So string theory continues to be a demonstrably self-consistent, (in principle) falsifiable, and stunningly beautiful theory--and dismissing it out of hand is a tragically common mistake.

Its worth mentioning that string theory went through quite few 'recalibrations', it feels like it string theorists are throwing stuff to see what sticks sometimes.

Could such a statement be tested?

What are the parts of string theory that are unique to it and not just invariants that apply to a large basket of theories?

Are there other theories which meet the same constraints?

Yes, and characterizing that common constraint seems more likely to be a fruitful path than assuming this coincidence validates string theory.

You can throw all sorts of interaction terms with unknown coefficients in there and the renormalization will pare down any inconsistent with large distance symmetries. The remaining unknown coefficients are equated to combinations of known and unknown physical constants, to find the unknown ones requires a new measurement (AKA massive international experiments or cosmological observation).

edit: Weinberg explains: https://www.int.washington.edu/PROGRAMS/talent13/refs/weinbe...

It has made some nice post-dictions: things spring out of that we do see, things not built into the model to start, which also lends credence.

Without understanding any of the details whatsoever: if it does provide a simpler explanation of known phenomena only, shouldn't it then be adopted by Occam's Razor?

The maths isn't rigorous and well-defined AFAICT. There aren't any definitive equations either, as far as I know. However, there are some cool geometric observations like AdS/CFT.

There are some very polemical critics of string theory like Peter Woit and Sabine Hossenfelder, who suggest that it's a complicated philosophy that started off simple and neat, encountered some problems, and then grew in complexity and vagueness.

Perfect for long careers built on publication output.

> math works out very nicely and it is so beautifully self consistent

AFAIK, nobody knows if it's self consistent yet.

Beauty is on the eye of the beholder. But I count "unsolvable" as a very ugly trait.

The entire point of string theory not being testable is not specific to string theory but rather to quantum gravity. Any theory of quantum gravity will only deviate from our current understanding (general relativity) at the Planck Scale. So, if one has an issue with string theory not being testable, one actually has an issue with pursuing any version of quantum gravity.

The interesting thing about string theory that sets it apart from all other theories of quantum gravity is that it's actually not just a theory of gravity. Because of this, there is actually hope that string theory could produce nontrivial testable predictions at energy scales much lower than the Planck Scale. However, it turns out that teasing out these predictions is very nontrivial, primarily because of a structure called the string landscape. Basically it turns out that the equations of string theory, while being tidy at high energy, produce a multitude of solutions at low energy and, unless you know which solution to look at, making definitive predictions is hard.

Normally the way around a situation like that is to make some observations and try to constrain which solution we should look at. Unfortunately our current best understanding of the landscape is only one-way; we can't currently constrain which solution to look at based on observational evidence. It's sort of like a one-way hash. If we can invert this 'hash function', we could in principle make measurements to constrain which solution we're in and at that point string theory would be falsifiable.

This isn't how working physicists talk, though. The physics researchers who develop mathematical models for physics are called theorists.

A string hypothesis would imply that they've come up with anything that's actually testable.

I know people attempted to correct the definition of 'theory' to discredit the 'It's only a theory' argument, but claiming it's only called a theory when it's "proven" is an overcorrection. In fact no theories are "proven" per se, the better theories just have known limitations.

More like a String Idea.