Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Fitting an elephant with four non-zero parameters

arxiv.org

Post reply on HN
Print view View original

101–110 of 155 posts

Re: Fitting an elephant with four non-zero parameters

#101

Post by olalonde »

edflsafoiewq wrote: post #48
maitola wrote: post #16

I love the ironic side of the article. Perhaps they should add the reason for it, from Fermi's and Neumann's. When you are building a model of reality in Physics, If something doesn’t fit the experiments, you can’t just add a parameter (or more) variate it and fit the data. The model should have zero parameters, ideally, or the least possible, or, even at a more deeper level, the parameters should emerge naturally fr…

Isn't the form of an equation really just another sort of parameter?

Kolmogorov complexity[0] solves this loophole :)

[0] https://en.wikipedia.org/wiki/Kolmogorov_complexity

Re: Fitting an elephant with four non-zero parameters

#102

Post by Pamar »

ahazred8ta wrote: post #55
maitola wrote: post #16

I love the ironic side of the article. Perhaps they should add the reason for it, from Fermi's and Neumann's. When you are building a model of reality in Physics, If something doesn’t fit the experiments, you can’t just add a parameter (or more) variate it and fit the data. The model should have zero parameters, ideally, or the least possible, or, even at a more deeper level, the parameters should emerge naturally fr…

Notably done for the first time irl in "Least square fitting of an elephant", James Wei (1975) Chemtech

https://demonstrations.wolfram.com/FittingAnElephant/

Re: Fitting an elephant with four non-zero parameters

#103

Post by Xcode23 »

kgeist wrote: post #99
visarga wrote: post #15

Earlier quoted context omitted.

> Whether "just curve fitting" is enough to produce something "intelligent" is, IMO, currently unanswered Continual "curve fitting" to the real world can create intelligence. What is missing is not something inside the model. It's missing a mechanism to explore, search and expand its experience. Our current crop of LLMs ride on human experience, they have not largely participated in creating their own experiences. Th…

>It's missing a mechanism to explore, search and expand its experience. Can't we create an agent system which can search the internet and choose what data to train itself with?

you need to define what the utility function of the agent is so it can know what to actually use to train itself. If we knew that this whole debate about human intelligence in computers would either be solved already or well on its way to being solved.

Re: Fitting an elephant with four non-zero parameters

#104

Post by beeforpork »

GuB-42 wrote: post #64
karmakaze wrote: post #43

Earlier quoted context omitted.

That's how I feel about dark matter. Oh this galaxy is slower than this other similar one. The first one must have less dark matter then. What can't be fit by declaring the amount of dark matter that must be present fits the data? It's unfalsifiable, just because we haven't found it, doesn't mean it doesn't exist. Even worse than string/M-theory which at least has math.

The dark matter theory is falsifiable. Sure we can't see dark matter (it doesn't interact electromagnetically), but we can see its effects, and it has to follow the laws of physics as we understand them today. It is actually a satisfying theory with regard to the Occam razor. We don't have to change our laws of physics to explain the abnormal rotations of galaxy, we just need "stuff" that we can't see and yet interac…

'dark matter' is not a theory, it is the name of an observational problem.

There are many theories to explain dark matter observations. MOND is not a competitor with 'dark matter', because MOND is a theory and it tries to explain some aspects (spiral galaxy rotation) of what is observed as the dark matter problem, which consists of many more observations. There is no competition here. There are other theories to explain dark matter, like dark matter particle theories involving neutrinos or whatever, and these may be called competitors, but dark matter itself is not a theory, but a problem statement.

Re: Fitting an elephant with four non-zero parameters

#105

Post by nertirs »

soist wrote: post #73
HeatrayEnjoyer wrote: post #70

Earlier quoted context omitted.

On what priors are you making that statement?

Rephrase your question. I don't know what you're asking.

I think he meant to ask, what is the difference between an experience and a predefined instruction?

Re: Fitting an elephant with four non-zero parameters

#107

Post by jampekka »

nyssos wrote: post #93
adrian_b wrote: post #90

Earlier quoted context omitted.

"Seeing" is indeed a poorly chosen word. The problem with dark matter is that there does not exist any second relationship from which to verify its existence, like in the case of normal matter, which takes part in a variety of interactions that lead to measurable effects, which can be compared. The amount and the location of dark matter is computed from the gravitational forces that explain the observed movements of…

All major DM candidates also have multiple interactions: that's the WI in WIMP, for instance. In fact I don't know that anyone is seriously proposing that dark matter is just bare mass with no other properties - aside from the practical problems, that would be a pretty radical departure from the last century of particle physics.

No interactions have been found, despite a lot of resources put into the search. So currently all dark matter particle theories apart from "non-interacting" have been falsified. And non-interacting theories are probably unfalsifiable.

Radical departure may well be needed, for other reasons too.

Re: Fitting an elephant with four non-zero parameters

#108

Post by XorNot »

beeforpork wrote: post #104
GuB-42 wrote: post #64

Earlier quoted context omitted.

The dark matter theory is falsifiable. Sure we can't see dark matter (it doesn't interact electromagnetically), but we can see its effects, and it has to follow the laws of physics as we understand them today. It is actually a satisfying theory with regard to the Occam razor. We don't have to change our laws of physics to explain the abnormal rotations of galaxy, we just need "stuff" that we can't see and yet interac…

'dark matter' is not a theory, it is the name of an observational problem. There are many theories to explain dark matter observations. MOND is not a competitor with 'dark matter', because MOND is a theory and it tries to explain some aspects (spiral galaxy rotation) of what is observed as the dark matter problem, which consists of many more observations. There is no competition here. There are other theories to expl…

Yes and no...MOND's core proposition is that dark matter doesn't exist, and instead modified gravity does.

Whereas you can have many proposals for what dark matter is, provided it is capable of being almost entirely only gravitationally interacting, and there's enough of it.

MOND has had the problem that depending which MOND you're talking about, it still doesn't explain all the dark matter (so now you're pulling free parameters on top of free parameters).

Re: Fitting an elephant with four non-zero parameters

#109

Post by XorNot »

nyssos wrote: post #69
karmakaze wrote: post #43

Earlier quoted context omitted.

That's how I feel about dark matter. Oh this galaxy is slower than this other similar one. The first one must have less dark matter then. What can't be fit by declaring the amount of dark matter that must be present fits the data? It's unfalsifiable, just because we haven't found it, doesn't mean it doesn't exist. Even worse than string/M-theory which at least has math.

> What can't be fit by declaring the amount of dark matter that must be present fits the data? Tons of things - just like there are tons of things that can't be fit by declaring the amount of electromagnetically-interacting matter that must be present fits the data. You can fit anything you like by positing new and more complicated laws of physics, but that's not what's going on here. Dark matter is ordinary mass gra…

You also have to do other things like not break General Relativity.

Which MOND does: it creates huge problems fitting into GR.

Whereas dark matter as just regular mass that interacts poorly by other means does not.

Re: Fitting an elephant with four non-zero parameters

#110

Post by XorNot »

rocqua wrote: post #63
jampekka wrote: post #58

Earlier quoted context omitted.

The problem with dark matter is that there's no (working) theory on how the dark matter is distributed. It's really easy to "explain" gravitational effects if you can postulate extra mass ad-hoc to fit the observations.

If there are two different types of observations, and one parameter can explain both, that is pretty strong evidence. Put differently, dark matter is falsifyable, and experiments have tried to falsify it without success. Besides the idea 'not all mass can be seen optically' is not that surprising. The many theories on what that mass might be are all speculation, but they are treated as such.

It's worth noting that one dark matter explanation is just: it's cold matter we just can't see through telescopes. Or black holes without accretion disks.

Both of these are pretty much ruled out though: you can't plausibly add enough brown dwarfs, and if it's black holes then you should see more lensing events towards nearby stars given how many you'd need.

But they're both concrete predictions which are falsifiable (or boundable such that they can't be the dominant contributors).

Post reply on HN