10 ms·
How much of the world is it possible to model?
- mitchbob 3y agohttps://web.archive.org/web/20240115144427/https://www.newyorker.com/culture/annals-of-inquiry/how-much-of-the-world-is-it-possible-to-model https://web.archive.org/web/20240115144427/https://www.newyo...
- codeulike 3y agoLike model cars, model airplanes, and model trains, mathematical models aren’t the real thing—they’re simplified representations that get the salient parts right. Once a year the New Yorker rings up Dan Rockmore and asks him to try and explain something to do with mathematics to people who have tweed mousepads and a Haruki Murakami screensaver.
- anigbrowl 3y agoHaruki Murakami screensaver Sticking knives in your eyes, sitting at the bottom of a dark well, and completely trashing your place of residence but leaving a few beers untouched all work well.
- FrustratedMonky 3y agoCan you guys explain? I've read a lot of Murakami and not familiar with the screen saver joke.
- anigbrowl 3y agoI'm just riffing about the sort of things that happen to Murakami's lead characters, which would not be most people's idea of a good time. The first and third are references to Hard Boiled Wonderland and the End of the World, and the second to The Wind-up Bird Chronicle.
- 082349872349872 3y agoMy first reaction was: "all of it, if one doesn't care if the model be accurate", and then I realised that for many thousands of years the most popular models not only purported to explain all of the world but even many worlds beyond it: realms of the gods, various afterlives, etc.
- Koshkin 3y agoTrue. There’s another aspect to this, though: for example, quantum theory lies at the foundation of everything that we can experience, so, in principle, as far as we humans are concerned, and beyond, it should, in theory, be able to explain (to “model”) as much; yet, in practice, it cannot. On a large scale, reductionism does not work, while any theory that aims to explain “everything” is necessarily reductionist in one way or another, and as such it becomes so “inaccurate” as to be useless (despite being universal and true).
- Frost1x 3y agoIt’s actually one of the biggest lies we teach children through education. We put this little carrot of reductionist science in front of them that says we can explain reality and go from incredibly reductionist ideas to increasingly less abstraction as their ability to compute and reason with these models inconsistency with reality begins to become more obvious. Just ignore friction, pretend it’s in a vacuum, etc. They’re not bold faced lies because it’s really the best we have about consistently explaining reality but I think we should let kids understand that even the latest and greatest reduction theories of reality won’t unlock truth everywhere, just allows you to peek in at some places and perform some degree of interpretation and prediction. It’s still quite valuable but we need to stop telling everyone we have the answers because we don’t. We almost masquerade around like a religion to some degree and people use that in the same way, “well science says” or “according to this data and paper” —- well yea, maybe, but also probably not.
- FrustratedMonky 3y agoWhat is up with people throwing out science these days? How did we go from: Our simplified model Ignored friction. to We have no idea what is real, throw away all technology and lets live in the trees.
- gaiagraphia 3y ago"That's another thing we've learned from your Nation," said Mein Herr, "map-making. But we've carried it much further than you. What do you consider the largest map that would be really useful?" "About six inches to the mile." ""Only six inches!"exclaimed Mein Herr. "We very soon got to six yards to the mile. Then we tried a hundred yards to the mile. And then came the grandest idea of all! We actually made a map of the country, on the scale of a mile to the mile!" "Have you used it much?" I enquired. "It has never been spread out, yet," said Mein Herr: "the farmers objected: they said it would cover the whole country, and shut out the sunlight! So we now use the country itself, as its own map, and I assure you it does nearly as well." (From Sylvie and Bruno Concluded by Lewis Carroll, first published in 1893.) I wonder whether the above is applicable to modeling the world and its behaviour in other ways; fully understanding everything of the complexity of the world would require the power of the world itself.
- derbOac 3y agoThis has been on my mind the last few years and I've thought about writing an academic paper about it in one domain. I've sort of increasingly suspected that some domains might involve an intrinsic amount of stochasticity or uncertainty that is baked into things. As in, an amount that becomes nonignorable relative to what you're trying to predict. There's some math and comp sci papers that basically argue (in a proof sense) you there must be a limit to modeling. It's been awhile since I read the papers but my recollection is that the modeling process itself creates (necessitates?) a certain gap between the model and the real world that creates hard boundaries. For example, if the modeling step requires an amount of time A, some information is lost in that time. These are a couple of the papers: https://arxiv.org/abs/0708.1362 https://arxiv.org/abs/0708.1362 https://www.jstor.org/stable/10.4169/amer.math.monthly.121.06.498 https://www.jstor.org/stable/10.4169/amer.math.monthly.121.0... I suspect as the system you're trying to model is more complex or has more dependencies it gets worse.
- di4na 3y agoCan I recommend looking at what people studying Experts at work and how they manage to do better than they "should" be able to have found? There are tons of domains looking at this empirically :)
- DarkmSparks 3y agoQ:How much of the world is it possible to model? A:All of it. The only problem with living so far in the future is when you get back to reality its all repeats. The doctor strange scene in avengers was pretty realistic for those who get access to the tech https://m.youtube.com/watch?v=eGKPfZTXHsc&pp=ygUaYXZlbmdlcnMgcG9zc2libGUgb3V0Y29tZXM%3D https://m.youtube.com/watch?v=eGKPfZTXHsc&pp=ygUaYXZlbmdlcnM...
- jongjong 3y agoAll simulations of nature are grotesque approximations. Butterfly effect is real and it makes a mockery of complex models.
- Animats 3y agoSome classes of models diverge rapidly on noise. When I was working on ragdoll physics in the 1990s, I discovered that a one-bit change in a 64-bit floating point joint angle was enough that a ragdoll falling downstairs simulation diverged visibly within one second.
- xscott 3y agoIt makes a mockery of simple models too. Models of double pendulums and three body problems don't match the empirical world for very long.
- jongjong 3y agoThat's the thing. It seems simple at a glance but even in those cases, there is a lot more going on than we care to admit. Now try to simulate the entire planet and you end up missing a ridiculous amount of variables and ignoring complex feedback loops.
- mvdwoord 3y agoTry explaining this to the climate people...
- mihaic 3y agoThis article doesn't actually seem to explore too much the deep origins and cultural impact of this mode of modeling. Lewis Mumford for instance in The Myth of the Machine all the way in 1970 wrote much better about the limitations of this way of thinking, that didn't do the heavy philosophizing of Baudrillard. The fact that models are necessary constructs, but are merely heuristics was one of Kant's main riffs (as much as I could understand), so maybe we've actually gotten worse at perceiving the limitations of models since we've lost touch with the real world over the last decades.
- huppeldepup 3y agoI'm surprised neither the article nor the comments here mention Adam Curtis, especially his "All watched over by machines of loving grace".
- ossicones 3y agoThe excellent new book, "Modeling Social Behavior" by Paul Smaldino, starts with a quote from Jakob von Uexküll that I've abbreviated a bit here: "An unbroken description of reality would be simultaneously the truest and most useless thing in the world, and it would certainly not be science. If we want to make reality and therefore truth useful to science, we must do violence to reality. [...] In nature, everything is equally essential. By seeking out the relationships that seem essential to us, we order the material in a surveyable way at the same time. Then we are doing science." I appreciated how this quote emphasizes that science and modeling are inextricably connected.
- randcraw 3y agoThe limits of math-based modeling inevitably will also apply to AI & ML-based systems. The difficulty of establishing a baseline set of conditions on which to math model a non-linear process (i.e. non-convex) surely will be just as problematic when the baseline data behind a AI learning-based model must also change. The Achilles heel in both cases of modeling is our awareness that the model's assumptions go bad with the passage of time. The math-based decomposition of non-convex processes into coverage using multiple convex models presumes that the span of the full model is appropriately segmented and conditioned. Its sanity check is to reassess those segments and conditions periodically. But do we have comparable means of revalidating ML-based (induction) models, given our lack of understanding of their black-box innards? Given that AI models can and will fail in time due to model AND data disintegration, must we simply wait until each AI model's prediction accuracy goes to hell to then build an all-new ML model and renewed dataset -- experimentally starting over from scratch each time because we never understood the limits of the AI model and its dependency on its training data to begin with?