[Rayhem]

> I have no doubt quantum physicists know what they are talking about but...I always think it is the kind of excuse a schoolkid would give their teachers for their calculations being wrong.

Just to emphasize how extreme this dichotomy is, not only is quantum mechanics correct (in that it's a predictive model), it's the most correct physical theory humans have ever devised in that the measurements there have more significant figures than anything else.

[mcmoor]

It's interesting that semiconductor engineers have to directly wrestle with the magic that's quantum tunneling. This theory is really not just a theory.

[tonyarkles]

In some devices it's something they have to fight against. In others, like SSDs, it's the feature that makes them work! It's not just a theory and it's not just another flaw that we've got to work around, we've taken it from a theory and turned it into useful technology.

[neuromanser]

You mean it's not just a hypothesis.

I would really, really expect users of this site to know the difference.

[andrewflnr]

Users of this site who pay attention to how science is actually practiced, and not the oversimplified cartoon version they teach us in school, know that the boundary between a "theory" and a "hypothesis" is rather blurry. For instance, have you ever noticed how it's called "string theory" despite having no real evidence for it? Have you ever heard anyone, let alone a real scientist, complain about this nomenclature?

Early theories and fleshed-out hypotheses overlap a lot. There's no sharp transition from one to the other.

[snowwrestler]

String theory is supported by a ton of evidence in that it can produce many predictions and hypotheses that match observed data. That’s why it’s called a “theory” and why people continue to study it.

What’s missing is a test that would produce evidence that would allow us to distinguish between string theory and competing theories. But that’s not nearly the same thing as saying it has “no real evidence.”

[andrewflnr]

In fact that's exactly what I meant by saying it has "no real evidence". Otherwise you could claim the totality of the universe as evidence for your pet Theory of Everything as long as it's not actually falsified yet, including but not limited to a theory where fundamental particles are fairies who use slide rules to decide how to interact.

[volemo]

If you fairy theory predicts the outcomes as well as the other theories that are on the table now then indeed it belongs there with them until someone finds a way to distinguish them and prove one is a better model of the reality. Better yet if the fairy model makes the calculations "cleaner" and easier.

[philipswood]

Correct me if I'm wrong, but string theory is still in the "it's so pretty and elegant is must be true kind of territory".

Not only do they have no proof, most of the potential experimental confirmations are at energies so high they're effectively out of reach for the foreseeable near future.

[snowwrestler]

You're not wrong, but what you're not appreciating is that the criteria for "pretty and elegant" are the same for string theory, the standard model, general relativity, and any other physical theory: a relatively simple and internally consistent mathematical framework that, when carried through in calculations, produces predictions that match a wide variety of observations, and is not contradicted by any known observations.

Proponents of string theory have not been able to propose an experiment that would allow them to exclude other theories, thereby demonstrating that string theory is better. But by the exact same token, critics of string theory have not been able to conduct an experiment that contradicts string theory, thereby allowing them to exclude it. And science moves forward by excluding theories with evidence (not just complaining about them).

Discussions of string theory among physicists are deeply intertwined with concerns about who gets famous, who gets grants, who gets tenure, who gets endowed chairs, who gets on TV and sells books, etc. But these types of concerns are a constant background noise to the practice of science, going back hundreds of years. Every scientist on Earth tends a private list of the wrong people who are getting too many resources to study the wrong thing.

If the critics of string theory could prove it was wrong, they would have, but they haven't yet. That makes it provisionally correct. Not correct, necessarily--it could be wrong! But it's not wrong yet, which is better than quite a lot of scientific theories proposed across the span of history (for now...).

[AppleBananaPie]

I used to think this same thing but I went down a rabbit hole a few months ago listening to people very critical of string theory like angela collier, sabbine, eric weinstein, peter woit and some others on youtube. Yeah they all have their own quirks but after listening to their take on the history of string theory the common things are it hasn't produced anything but the major proponents of it always talk like it has and in some cases outright lie about things that it has contributed.

I really like listening to brian greene and sean carroll but now when I listen to them, particularly in recent videos, it feels like there is much less substance to what they're actually saying string theory has done.

But who knows! Maybe I'll learn something new and completely flip my world view again, I'm not a physicist by any stretch so have to rely on listening to experts :)

[infamouscow]

I think it's important to draw a distinction between the formal and natural sciences. HN is overrepresented by folks with backgrounds in mathematics and computer science where using "theory" is correct, e.g., set theory.

Bret Weinstein explained the distinction between a hypothesis and theory a few weeks ago on YOUR WELCOME.

https://www.youtube.com/watch?v=ac4MnNrs6g4

[andrewflnr]

I specifically cited an example in the natural sciences, specifically physics. IIRC chemistry and biology do similar shenanigans at times. Again, consider how you could possibly draw a sharp distinction when accumulation of evidence is so often a gradual process.

I'll consider responding to the video if you give me a reasonable timeslice, but I'm not going to watch the whole hour.

[elpocko]

What if they don't? Will they cease to be users of this site? What if they speak a lot more languages than you, just not perfectly? Does it even matter what you would expect?

[mcmoor]

The (scientific) theory is really not just a (colloquial) theory

[julianeon]

The other famous example is PET scanners, which actually use a form of antimatter: positrons (the antimatter counterpart of electrons).

[hwillis]

You don't even need that. The prediction he's referring to as most accurate is the magnetic moment of the electron, which is used in plain old MRIs. If we didn't have the quantum mechanical correction, all of our MRI images would be distorted. Only by a few pixels, but it'd be noticeable!

[sph]

We used fire in our daily lives even before organic chemistry existed.

[seeknotfind]

So many significant digits includes a level of self-consistency of the model, since we are assuming the model to some degree in order to measure it. Though in this case, it's not the calculations that are wrong, but the model, we hope is wrong. That is, a new perspective and a new way of thinking about things may reveal more. Of course, we are always fighting against the irreducible complexity camp. However, the fundamental lack of cohesion between quantum and relativistic theories demonstrates there is at least one big thing we are still doing wrong.

[hwillis]

> So many significant digits includes a level of self-consistency of the model

No, that's incorrect. The specific measurement -the most accurate scientific prediction humans have ever made, the anomalous magnetic moment- is only to 1 in 10 trillion. The magnetic moment (think moment of inertia) is the ratio of force from the magnetic field to the mass of the electron. You put an electron into a magnetic field, and it'll turn to face the field at a certain speed. If you stick the electron in a vacuum, it overshoots (because of rotational momentum) and ends up wobbling back and forth at a specific frequency. That's how MRIs work; they make a big magnetic field (stronger on one end) and then measure how many electrons are wobbling in different areas, since the electrons in high-strength fields wobble faster.

Specifically, you expect the wobble to get ~2.8025 GHz (similar to a microwave oven) faster for every 1000 gauss (please, no jokes about teslas). It's very convenient to measure a difference in frequency, since you can just measure the drift over time. Because that frequency is relatively high it takes about 30 minutes for the frequency to drift off by a half-cycle and totally cancel out your reference.

And it's super easy to get a reference frequency, too. It's just the charge of an electron divided by 2x the mass of an electron. Did you do this experiment in physics class? https://www.youtube.com/watch?v=Kcn2VgBNJjg

Then you measured the charge/mass ratio of an electron. A clock that's accurate to 1 in 10 trillion is also not a big deal, although unless you have an oscilloscope in your house its probably more accurate than any clock you own. Still, you can buy a better Phase Locked Loop for a couple bucks.

If you just wanted to measure the difference, you don't need that much precision, though. The correction from quantum mechanics is pretty large, relatively speaking: ~0.16%. Even the next several digits are super easy, and it's only those last ones that you really need to bust out the liquid helium.

Lawrence and Livingston made the first cyclotron out of literal junk: https://upload.wikimedia.org/wikipedia/commons/6/61/4-inch-c...

And it really doesn't take much more than junk to get to that 0.16% accuracy that lets you see that a classical prediction of the electron is just VERY wrong. But if you listen to those wizards with the bongos, suddenly they're really, really right about what that difference is: https://en.wikipedia.org/wiki/Vertex_function

[seeknotfind]

For instance, there is disagreement on if time is quantized, and relatively says time is relative. So at the point we might reimagine the basis of our understanding, two independent measurements of Hz are incomparable. Whatever is out there for us to discover, I'm sure it'll take us all by surprise. I've no doubt you know way more about how we measure these things in practice than I do, but our scientific model and measurements are inextricably linked. This isn't a useful way to think for most practitioners, but it's the perspective that the next Newton or the next Einstein will need to consider. So, in rootier poster comment that this feels like a wrong calculation, I want to treasure that part of the perspective that seeks to reimagine the fundamentals.

[hwillis]

> but our scientific model and measurements are inextricably linked. [...] This isn't a useful way to think for most practitioners, but it's the perspective that the next Newton or the next Einstein will need to consider.

I totally get it, and you aren't wrong, but this is literally the thing it applies to the least. The core thing here is that the obvious assumption about how the universe would work is very (>.1%) wrong, and the purely theoretical quantum math is incredibly, absurdly, amazingly correct.

It's not about measuring some amazingly small thing and having it be amazingly correct. It's about us seeing an incredibly large error in how the universe seems to behave, which QED explained perfectly. It's like the ultraviolet catastrophe or einstein's cross except its billions of times more correct.

[blackhaj7]

That’s awesome to hear.

I totally back the scientists and wish I could understand it better but I always like to have a chuckle that the crazy sounding parts are just the scientists making up stuff

[ronald_raygun]

Just because it’s quantitatively accurate doesn’t mean it’s “true”. Like a fun fact I like is the geocentrism was extremely accurate in terms of astronomical predictions when heliocentricism came about (it was actually more accurate for a while).

[alok-g]

How do you define 'true'?

If two theories are working very well with no experiment available to indicate right vs. wrong for them, Occam's Razor would be all we would have to make a choice. That however does not really make one of them truer than the other.