Hacker News Reader

Comment by lazide | Hacker News Reader

[flagged]

"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."

"Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith."

https://news.ycombinator.com/newsguidelines.html

Hardly a shallow dismissal dang, as the replies show. It’s a very valid critique of where it has ended up, and goes right to the heart of the underlying problem.

The challenge is if it’s intentional from the start, or merely ended up being intentional at the end eh?

And the folks arguing that the underlying critique are false, as shown in the follow-ups, are wrong.

If anything, it’s just getting downvotes because people don’t realize how on the nose it actually is, near as I can tell.

Sorry, but your GP comment consisted of nothing but putdowns—not just of an entire field but of the people working in it. That is a classic shallow dismissal in the sense that we use the term. Not a borderline call!

I guess we have a different definition of put downs eh? Noted, thanks.

I think that you have half a point. You're absolutely right that just because people are paid to think about things, doesn't mean that they are making progress. And there is a lot of evidence that this is true today in the foundations of physics.

However string theory was not intentionally untestable. In https://www.youtube.com/watch?v=eRzQDyw5C3M she gives a good history of why it was originally invented, what testable predictions it made, how it failed those tests. And then how string theorists who were trying to find relevance for their work tried to keep it going as it stumbled into being untestable.

Fair point, I guess. It’s easy to also see how it just mostly ended up there. Still. A problem.

This is a very cruel reading of string theory. Intentional? What?

Exactly. Consistently untestable and unfalsifiable claims for decades has to be seriously questioned at some point, and I think we're well beyond that point. This is especially true for string theory. I'm particularly fond of how Angela Collier laid out the timeline of string theory in her video on it[0] as well as the consequences that science communication is now facing as a result.

[0] = https://www.youtube.com/watch?v=kya_LXa_y1E

The same could've been said of atomic theory, neutrinos, gravitational waves, the higgs boson, cmb radiation, plate tectonics, and quantum mechanics at various points in time.

Weren't those all arrived at from a series of falsifiable predictions? What does string theory even predict that can be tested?

As a non string theorist my understanding was that string theory actually makes quite a lot of empirically verifiable statements, just that those statements are only interesting at either never or extremely high energies.

I think ppl are asuming that sting theory comes from the meme about turning 1+1 = 2 into some massive integro differential equation. The world is rarely so simple.

I’ve heard that it also predicts at very low precision, some values that are practically measurable, and, unsurprisingly for how little precision these predictions have, these predictions are correct (I.e. the experimental results are within the predicted range).

(Or, maybe “a prediction” rather than “predictions”? I only heard about one, and I forget what it was.)

I am aware of no case where it clearly made an advance prediction of any behavior that later turned out to be true.

I'm aware of quite a few where they managed to "predict" something we already knew.

That said, they've made so many "predictions" that I'm sure that some likely worked out by sheer coincidence.

Oh, yes, I meant predict a value we had already measured at the time the "prediction" was made. I should have made that clear in my original comment. I would add it now except that the editing time has run out. Maybe I should have said "postdicted".

Actually, I think the value might have been something like, the electron mass? Or something like that. (Which, obviously, had been measured before string theory made a "prediction" of it.)

I think the prediction you may be referring to is supersymmetry, which was apparently empirically disproved by the LHC, or at least the supersymmetric extension to the standard model was disproved.

https://www.sciencenews.org/article/supersymmetrys-absence-l...

You are making it sound as though string theorists are asserting some kind of flying spaghetti monster theory. Do you think these people are not genuinely interested in advancing science? That's an ad hom fallacy. There is a difference between a hypothesis being conceptually unfalsifiable and a hypothesis that is incredibly difficult to test from a practical standpoint, or impossible with present energy constraints.

I don’t think the mistake made is exactly an ad hom fallacy? I agree with the rest of your comment though.

That statement is only true for a few of the things on your list..

Yes, it took a couple of decades to test the existence of neutrinos. But, for example, general relativity was successfully tested within 5 years of being published. Gravitational waves were a prediction that took decades before we could test them, but the theory itself had lots of other verifications.

To date string theory has had many predictions that leads to failed tests. But not a single successful test in its favor.

No, not really. All of those had reasonable, technically addressable methods for testing.

No you couldn’t. And it’s been 80 years now!!

All of those things you name came directly out of attempts to create testable hypotheses from experimental observations, and all of them were tested as soon as anyone could build an experiment apparatus or gather the data to do it. Which didn’t take that long considering the extreme engineering difficulties in actually building the apparatus for some of them.

String theory has avoided testability it’s entire existence, nearly a century now, and no one that I’ve seen is even attempting to make an experiment to try to test it - because at this point it’s clear that no one on the theory side is interested in making a testable hypothesis. That isn’t luck, that’s talent and hard work.

It’s one of the most absurd grifts I’ve personally seen play out so far.

80 years? I would date its birth as 1968-9 (Veneziano), it’s hard for me to imagine calling prior work than that as “string theory”. But never mind that—the bigger problem with this (quite common) argument is that everything about quantum gravity, not just string theory, has avoided testability because our other theories are too good, and because we’re limited to doing experiments on Earth with equipment built on human scales with human budgets, and that’s just not where quantum gravity would naturally make itself known. So really this argument just suggests we shouldn’t study quantum gravity at all. Maybe that’s your actual opinion—it’s a waste of time if we can’t access the Planck scale, we should table it all and sit on our hands until we can. But string theory really is quite interesting to study, stuff like AdS/CFT is just really surprising and magical when you get what it’s about, and it would be a real pity to not pay the meager salaries of theoretical physics just because of pessimism. String theory is so far from fully understood! It’s actually…really hard!

BtW I think you got this 80 years number from looking at the earliest date on the Wikipedia page. You might want to read it more carefully. Not everything leading up to string theory is string theory.

Fair enough - 50 to almost 60, not counting s-field precursor work.

I’m not saying string theory isn’t potentially interesting from a mathematics perspective, I’m just saying treating it like physics (which is, explicitly about testable/falsifiable theories) is BS.

If we were honest about it, it would be a maths speciality eh?

At least until there are more clear attempts at making testable hypotheses.

But that would cause other issues with funding I imagine.

If quantum loop gravity comes up with a testable hypotheses, then hey, maybe I’m wrong. But so far, not so much yeah? And I’m not talking ‘we’d need to spend a lot of money to test it’, I mean an actual testable/falsifiable hypotheses at all.

There is no evidence to suggest that string theorists designed the theory to be untestable.

After this much time and that much work, how is it possible for a physics theory to not have a single testable/falsifiable prediction without it being intentional? It has been over 80 years. [https://en.m.wikipedia.org/wiki/History_of_string_theory].

The evidence is in the absence.

It predicted supersymmetry, which has been experimentally disproved.

The kind of supersymmetry you’re referring to (global spacetime supersymmetry) is not required by string theory; this is a common misconception. Looking for super partners in a collider is actually only telling you about global supersymmetry, which unlike local supersymmetry is not a universal feature of string theory at low energy, in fact the opposite, it is probably non-generic. It so happens that a class of appealingly simple vacua do have this property, which led to some inappropriate optimism among string theorists that has entirely abated with more experiments. Unfortunately this has been widely misunderstood to rule out the whole enterprise of string theory, which is unreasonable for the reason stated above, it is much more likely to not see SUSY below the Planck scale. [0] (Unless you just like to mock string theorists for hoping that the universe would be kind to them.)

Also global supersymmetry has not been experimentally disproved (how would you do this, even?) but it is true that current or even near-term experiments are not nearly sensitive enough to get close enough to answering this definitively, which is obviously upsetting.

[0] https://ncatlab.org/nlab/show/string+theory+FAQ#DoesSTPredic...

Not intending to mock anyone and I don't know nearly enough physics to have a credible opinion either way. Thanks for your explanation.

Don’t worry, I didn’t think you were :) you’re welcome!

That proves it!

I mean, you should prove they didnt. If that sounds unreasonable we've made progress. Prove we didn't?

Ill let myself out

How is it "intentionally untestable"? I get that it is practically untestable, but as far as I know, there are people working to try to find some possible tests.

Reply on news.ycombinator.com