AI Cracks Centuries-Old 'Three Body Problem' In Under a Second (livescience.com) 146
Long-time Slashdot reader taiwanjohn shared this article from Live Science:
The mind-bending calculations required to predict how three heavenly bodies orbit each other have baffled physicists since the time of Sir Isaac Newton. Now artificial intelligence (A.I.) has shown that it can solve the problem in a fraction of the time required by previous approaches.
Newton was the first to formulate the problem in the 17th century, but finding a simple way to solve it has proved incredibly difficult. The gravitational interactions between three celestial objects like planets, stars and moons result in a chaotic system -- one that is complex and highly sensitive to the starting positions of each body. Current approaches to solving these problems involve using software that can take weeks or even months to complete calculations. So researchers decided to see if a neural network -- a type of pattern recognizing A.I. that loosely mimics how the brain works -- could do better.
The algorithm they built provided accurate solutions up to 100 million times faster than the most advanced software program, known as Brutus. That could prove invaluable to astronomers trying to understand things like the behavior of star clusters and the broader evolution of the universe, said Chris Foley, a biostatistician at the University of Cambridge and co-author of a paper to the arXiv database, which has yet to be peer-reviewed.
Newton was the first to formulate the problem in the 17th century, but finding a simple way to solve it has proved incredibly difficult. The gravitational interactions between three celestial objects like planets, stars and moons result in a chaotic system -- one that is complex and highly sensitive to the starting positions of each body. Current approaches to solving these problems involve using software that can take weeks or even months to complete calculations. So researchers decided to see if a neural network -- a type of pattern recognizing A.I. that loosely mimics how the brain works -- could do better.
The algorithm they built provided accurate solutions up to 100 million times faster than the most advanced software program, known as Brutus. That could prove invaluable to astronomers trying to understand things like the behavior of star clusters and the broader evolution of the universe, said Chris Foley, a biostatistician at the University of Cambridge and co-author of a paper to the arXiv database, which has yet to be peer-reviewed.
But it still can't drive a car without human help (Score:1)
Re: (Score:3, Interesting)
Re: But it still can't drive a car without human (Score:3)
Black holes (Score:2)
Re: Black holes (Score:2)
Re: (Score:2)
Re: (Score:2)
a fact which is also true for many humans as well it would seem
Brutus or the New Algorithm: WHICH IS CORRECT?!? (Score:4, Interesting)
The algorithm they built provided accurate solutions up to 100 million times faster than the most advanced software program, known as Brutus.
So they ran Brutus for several months to get what they believe to be the 'correct' answer, and then the new algorithm accomplished something-or-other else in an elapsed time of, say (three months) / (100,000,000).
Lets call that 90 days, and assume 60 seconds in a minute, 60 minutes in an hour, and 24 hours in a day.
Then we're looking at an elapsed time of:
Two points:
1) Unless they had access to something like the IBM Summit supercomputer, and somehow they seeded the various aspects of the parallelization algorithm to each of the nodes of Summit before shouting "Go!", I'm calling bu11sh!t on any meaningful work being accomplished in under 8/100ths of a second.
2) Even using something as grotesquely inaccurate as 64-bit floating numbers [I'm gonna assume that they didn't divvy up the $$$s necessary to run this on 128-bit math hardware], presumably Brutus and this New Algorithm didn't come to precisely the same answer - each would have been off by a few bits from one another.
So we're dealing with three different quantities:
But this is CHAOS, ergo [ipso facto] we don't even know which of Brutus or the New Algorithm is producing the more accurate answer, much less which will diverge from the correct answer more rapidly.
PS: I'm not saying that these guys aren't onto something, but whatever they're onto probably isn't gonna be classical DEDUCTIVISTIC mathematics - it's probably gonna be some new EMPIRICAL black-magic voodoo mathematics [which works when it works, but when it fails, could very well fail spectacularly].
And I don't wanna leave the impression that I would necessarily be opposed to black-magic voodoo mathematics - it could be a fascinating new endeavor for our soon-to-arrive CRISPR'ed Gattaca Super-Babies to ponder [if they can ever pull themselves away from the lure of the 24x7 (((pr0n))) being streamed to them on their 3D contact lenses].
And I probably shouldn't joke about 24x7 (((pr0n))) being streamed on 3D contact lenses - that might be here by 2030 [if not sooner].
Re:Brutus or the New Algorithm: WHICH IS CORRECT?! (Score:5, Insightful)
1) You don't think you can have a meaningful calculation in 80 milliseconds? The actual paper suggests, by the way, that the time taken was on the order of 1 millisecond, no need to reverse engineer. It's a fixed-time result. Which means that I agree it can't be totally general in an analytic sense.
2) In physics the inputs are *always* precision-bounded, so outputs are always in comparison, even in chaotic systems. Chaos here is a relative term, not absolute. Almost any chaotic system is non-chaotic when the changes in initial conditions are constrained enough. Given the Heisenberg uncertainty principle, and well, practical reality, a truly absolutely chaotic system is one we cannot make predictions on, so it's ultimately indistinguishable from truly random and therefore irrelevant to something we think we can calculate. In any case, chaotic systems are sensitive to initial inputs, but the accumulated error is *not* initial inputs, it's outputs (and intermediate internal states). The arxiv article contains analysis of error accumulation.
Re: (Score:2)
I once worked on a system that took 20ms to provide answers, and that involved running jobs on a small cluster. It was intended to replace existing systems that take about 6-12 months to process.
Yes, its answers were accurate and precise. The algorithm was similar in concept to adding an index to a database table. This is exactly how major jumps in computer science occur: take an existing problem, rearrange its parameters into a different problem, and try to solve that one faster than the original.
Sometimes
Re: (Score:2)
In this case though, what they really developed is essentially a template system. The big number cruncher is still needed to build the patterns, and then the neural n
Re: (Score:2)
Part of my PhD was a O(N log N) algorithm to replace a O(N^2) one. It brought a computation we wanted to do down from about 2 million years to twenty or 30 milliseconds.
Re: (Score:2)
Also relevant, is that a neural network is effectively a type of lookup table. There is a lot of information pre-compiled into the weights of the completely trained neural network. This could explain how an NN approach can come up with results a lot faster than a naive brute-force computation.
Re: (Score:2)
Re: (Score:2)
Re: But it still can't drive a car without human h (Score:2)
A robot cat. And he can drive, just not very well.
https://giphy.com/gifs/snl-nbc... [giphy.com]
That's not cracking the problem (Score:5, Insightful)
That sounds like an extremely useful algorithm, but if I read the summary correctly, it's still just an approximation. It's not a solution.
Actually, after reading the abstract it sounds even more limited than that, though it's hard to say just what the limits are. "Our results provide evidence that, for computationally challenging regions of phase-space, a trained ANN can replace existing numerical solvers, enabling fast and scalable simulations of many-body systems to shed light on outstanding phenomena such as the formation of black-hole binary systems or the origin of the core collapse in dense star clusters. "
Re: (Score:2)
But do the scientists know the actual algorithm that the computer used or is it a black box? If the later then we're on the way to trusting our future to Skynet.
Re: (Score:2)
The state of the network *is* the actual algorithm, so yes, they know that. What they don't know is the logic by which the network arrived at that configuration.
Re:That's not cracking the problem (Score:4, Informative)
Exactly. Cracking the three body problem involves a closed form solution of the equations of motion. This is just a new computational method.
Re:That's not cracking the problem (Score:5, Interesting)
Your answer is probably correct. But since the 3-body problem is chaotic, if there's a localized hole or attractor [wikipedia.org] which wasn't present in the data used to train the ANN, then the results it gives will be completely wrong in that region of the solution space.
Re: (Score:2)
It sounds like there are limits other than the training data. It's hard to be certain, though.
In related news ... (Score:2)
Re: (Score:2)
In some fields even that would be a massive improvement. It's a computer, not a messiah. It's not going to break the laws of physics.
similar common subproblems (Score:2)
Re: (Score:2)
AIs will suck at extrapolation for the same reason Taylor series are useless for extrapolation -- because they're focusing entirely on fitting the existing data to the curve and interpolating. But maybe (like a Taylor series) the fitted region will provide a term in a larger series as the borders of the approximation get expanded, meaning AIs can build on the models discovered by other AIs, find out where they overlap and can be stitched together to provide a bit more generality, and then generate new model
Re: (Score:2)
Please stop implying machines have intelligence (Score:5, Insightful)
They do not. All we have is statistical classification, automation and some rule-based inference. There is absolutely no intelligence or insight in machines today and that will remain true for the foreseeable future.
Re: (Score:2)
Re: (Score:2)
Well, not really. Maybe calling it FI (Fake Intelligence) would be an appropriate description. Of course, anybody in the know has started real intelligence created artificially AGI (Artificial General Intelligence), but most people do not understand that.
Of course, if you look at "artificial xyz", it usually is not very good and quite often not a real replacement for the real thing, so you do have a point.
Re: (Score:2)
Re: (Score:2, Insightful)
Re: (Score:3)
Actually, computers cannot do reasoning or knowledge at all. Both things require understanding and insight. What they can do is data storage and real simplistic logical and arithmetic operations really fast.
Re: (Score:2)
There is absolutely no intelligence or insight in machines today and that will remain true for the foreseeable future.
You'e exactly right. However and somehow, they're still smarter than some people I know. Different domains of expertise, I guess.
Re: (Score:2)
Well, the thing about people is that while everybody does actually have intelligence, most people rarely use it and when they do, they do so only in a limited fashion and only applied to certain types of problems. For most things, most people just copy what others do without ever thinking about it. Then you have the about 10-15% "independent thinkers" and these people are able and willing to apply their intelligence to anything and there you can see what is actually possible. Unfortunately, being able to th
Re: (Score:2)
What we're really going to discover is how little is actually intelligence and how much is passed down skill and knowledge. Even the things that are original thought will often end up being things other smart or skilled people have thought of before that we're simply not aware of or that they've never managed to articulate to others. I wish I knew all a Michelin star chef knew about cooking but I don't, in fact any moderately trained AI is probably better than me because it'll mimic skills far beyond my own
My prediction (Score:2)
I haven't looked at the problem for ages, but if you are interested, there is a nice paper here: https://iopscience.iop.org/art... [iop.org] . It discusses calculation of positions of planets etc in the solar system over the last 200 million years. It's just _slightly_ more difficult than the 3-body problem. There are probably newer papers.
Re: (Score:2)
Re: (Score:2)
Very misleading (Score:2)
This is not "solving the 3-body problem" in any way that you would recognize. Solving it, in mathematical terms, would require a closed-form solution. This appears to just be a different way to do the integration of a numerical solution. That's not "solved". In any case, most practical applications of 3-body gravitational dynamics - like navigating to the moon with errors on the order of 100's of feet - was "solved" in the same sense back in the 60's in seconds, and could be solved in the same sense now in
It's been cracked before (Score:2)
Smart idea, dumb reporting (Score:2)
Once again, the reporting is overly flamboyant, but the actual article is quite nice.
The abstract is at: https://arxiv.org/abs/1910.072... [arxiv.org]
The article is at: https://arxiv.org/pdf/1910.072... [arxiv.org]
Science and mathematics have done a great job of understanding the interactions of two bodies, or else one body (parameter, variable, whatever) in time. But, 3 bodies or N-body interactions do not have discrete equations of state. The best you can do is take a system of multiple equations describing the 2-body intera
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
There’s literally no intelligence in artificial intelligence research.
The purpose of AI research is to build intelligence out of parts that are clearly not intelligent. Similar to how your brain is made from leftover parts from food. You are looking at the field, and observing the researchers using non-intelligent parts, and then conclude that the whole thing can't be intelligent.
Re: (Score:2)
So yeah, AI research overlaps with a lot of other fields, that is kinda how brains, intelligence, and reasoning work. A significant amount of cognition is essentially squishy inference engines, so it makes sense that researchers would draw from fields that are also exploring those kinds of mathematical models. Take that away and all you hav
Re: (Score:2)
Re: (Score:2)
Re:Not Artificial Intelligence (Score:5, Informative)
They dont seem to have produced a method of calculating the position of 3+ bodies at an arbitrary future time T. What they seem to have done is produce a better integrator, that can step along time quicker. Still an iterative algorithm, so still unsolved.
Re:Not Artificial Intelligence (Score:5, Insightful)
Exactly. TFA talks about a great new tool to iterate a solution quickly, but is is not what is meant by solving the N body problem. Such a solution would be parametric equation as you say and would work on any arbitrary N bodies.
Even worse.. (Score:4, Interesting)
Worse worse... all they seem to have done is made a system that allows for a lower error over a bit larger timestep, hence making it look 'faster', however it cannot give error bounds! so you have no way of tracking the cumulative error... on an iterative solution....
In other words, its junk for any real use - but at least its fast!
There are similar systems for turbulent fluid flow... they work very well... most of the time, until they dont.
Re: (Score:2)
Re: (Score:2)
I haven't read the whole paper yet, but I don't think that's the case. A deep learning network is a parametric model. They put in the initial conditions and the time they want and it spits out the positions at that time. They trained it using estimates from an iterative model.
It's technically not a full solution because it works over a bounded time interval, but that's being pretty pedantic.
Re: (Score:2)
It's technically not a full solution because it works over a bounded time interval, but that's being pretty pedantic.
Sorry bud, but if "working over a bounded time interval" is allowed, then there is no N-body problem (Euler integration, you know, integration from hundreds of years ago, also works over a bounded time interval) and thus these guys solved a non-problem.
In 2-body problems, there is no time bound.
Full stop and turn in your nerd card because you dont understand math.
Re: (Score:2)
You're aware Euler's method is iterative?
Sorry "bud", keep working on the math thing.
Re: (Score:2)
Pattern recognition is not artificial intelligence.
Regular expressions are patterns. Solving them doesn't even require a Turing Machine, so sure, that is not AI.
But looking at what is happening in a social situation and accurately predicting what will happen next is also pattern recognition, and is certainly AI.
This system is looking at a complicated physical system and predicting its behavior in a way that is analogous to intuition. Many people would consider that AI.
Re: (Score:2)
Actually, you're wrong about that. Put slashdot.com into your browser.
Re:Chaotic systems and humility (Score:5, Insightful)
Ahh yes, the appeal to complexity - fallacy of climate deniers and conspiracy theory crackpots the world over.
It's not complex at all. Carbon dioxide is a greenhouse gas. This is provable and very simple to understand. We have greatly, vastly increased the amount of CO2 in the atmosphere by digging up fossil fuels and burning them. This is provable and very simple to understand.
Put a) and b) together and you get climate change. By all means reasonable people can debate the immediacy/timeline and whether some people are being overly alarmist, or whether some are trying to profit from it. Sure, knock yourself out. But you and your ilk seem to be trying to 'debate' simple facts, which is never useful to anyone reasonable.
Re:Chaotic systems and humility (Score:4, Interesting)
And yet, the Climate Scientists have proven themselves incapable of putting out a falsifiable prediction, that didn't get falsified in due time...
Knock yourself out — try it. Provide a list of predictions, that came true. The list will consists of pairs of links — the first to the prediction being made, the second — confirming it coming true.
Other requirements are:
The above requirements are perfectly sensible — if "Climate Science" were a real scientific discipline, it would've produced numerous such predictions the way Physics or Biology do. Try it...
I can use this same line of reasoning to "proof", how humanity can push the planet off orbit. It's not complex at all. Any time a ball is thrown, an artiller shell fired, or a rocket is launched in one direction, the planet is nudged in the other. This is provable and very simple to understand. We have greatly, vastly increased the number of such throws, shots, and launches — because of our higher numbers and technological advances. This is provable and very simple to understand.
Put a) and b) together and you got the planet shaking in its orbit. By all means reasonable people can debate the immediacy/timeline and whether some people are being overly alarmist, or whether some are trying to profit from it. And so on up to and including the "your ilk" part...
Re:Chaotic systems and humility (Score:4, Insightful)
Climate scientists predicted that the temperatures would rise, sea levels would rise, and polar ice caps would shrink.
Temperatures have risen, sea levels have risen, and polar ice caps have shrunk.
Models can make useful predictions that nevertheless contain some error. The goal is to create models and continue to refiine them so that they have as little error as possible. That's not unmeaningful or impractical -- that's just how you build models.
Science makes predictions all the time that contain some error. Get over it.
Re:Chaotic systems and humility (Score:5, Informative)
Here's a webpage with a meta-analysis of old climate predictions and giving the difference between predictions and observations, all well beyond 5 years:
https://www.carbonbrief.org/an... [carbonbrief.org]
Not all of the ones that were looked at were within 20%, though all were within 30%, and not all of them were over or underestimates, and 20% is pretty much an arbitrary limit that isn't applied generally to all cases of other sciences like Physics and Biology. *Especially* when error bars are provided beforehand.
if "Climate Science" were a real scientific discipline, it would've produced numerous such predictions the way Physics or Biology do.
Funny you should say this, this *is* predictions using physics. Biology predictions are quite difficult and it's interesting you should say that. The obvious analogy is to predict the future course of evolution for a multicellular nimal species given predicted changes to the environment.
Also, your last argument is not sound, you do not get net linear motion from moving internal portions of something. Throwing an artillery shell just changes the shape of the Earth, but nothing leaves the Earth. Only things like space shuttle launches have any real relevance, and even that has a calculatable influence.
Re: Chaotic systems and humility (Score:2)
Complexity DENIALIST!!!!1!!
Re: (Score:2)
And it is you that is ignorant of the nature of chaotic systems, in which something -even so relatively simple- as three spheres leads to incalculable behavior, and are denying these "simple facts".
Again, gases are chaotic systems, yet despite our inability to predict the behavior of individual molecules, we can still make engines work based on statistical physics. What's so difficult to understand for you here?
I'll bite (Score:2)
One word: Trees.
More CO2, more tree grow sequestering the CO2.
Here's a graph of the extra CO2 [ourworldindata.org] rising from ~zero to 35 billion tons a year.
Can you show how the number of trees similarly increased?
Since say 1950 when it really took off. Where did all the extra trees spontaneously grow?
Re: (Score:1)
You seem hysterical. Natural way to confess your viewpoints' weaknesses, from the perspective of you yourself.
That aside, yes, trees. We plant more. That alters the scenario. And, note a large contingent of scientists has already recommended precisely that, before you hysterically dismiss that.
Are this and future analyses and remediation strategies to be excluded from the nature of the system over the future timeframe? Are you denying those unknown strategies are known to not be a factor in this chaoti
Re:Chaotic systems and humility (Score:5, Insightful)
So wait. Let me dissect this a bit. You aren't arguing that there are more trees and we don't need to worry, but that we should plant more trees? That would be..mitigating climate change, hence admitting it exists.
You're talking out of both sides of your mealy mouth, and the "look, I'm being calm here" approach is dreary and unimpressive. So now it's not 'chaos that will fix itself and which we can't possibly understand' it's 'we should plant some trees, by george!'.
If your position is 'Climate change exists, is largely caused by man, and we need to take approaches to stop it', then I'm not sure what we're arguing about. Sure, let's plant some trees, shut down shitty coal plants, build some nuclear power plants, and invest heavily in research into fusion. I'm on board.
Re:Chaotic systems and humility (Score:5, Insightful)
Yet, the future interaction of hundreds of "Climate Change" causal factors is "settled science".
You're right, I'm totally off base for thinking you to be a climate denier simply because you talk in the exact and specific language of the "I'm just asking questions!" climate denier.
Climate change is fairly easy to understand, and there are many solutions but 'whelp, it's complicated who knows!' is definitely not one of them.
Re: (Score:2, Informative)
You put doubt-quotes around climate change and settles science. Climate change and the factors causing it (humans) are not in doubt. The amount of degenerate weaseling you've done in this thread is pretty amusing. Do you find that actually works on anyone?
Now, had you come out and said that you have issues with certain blowhards and lunatics, e.g. crazy breathless tween girl from Sweden or wherever, then sure that's something reasonable people can debate. Instead you twist in the wind, backtrack, and make f
Re: (Score:2)
Nice proof link.
Unfortunately the degree of intellectual dishonesty of both posters and mods to direct facts that contradict their wishes, is ridiculous.
Fortunately, it will inevitably be corrected, with appropriate compensation.
Re: (Score:2)
The climate will change regardless of what we do. You are presenting/ranting on what a -correct amount- of climate change should be, per you.
The IPCC publishes, as part of their periodic reports, a glossary with the definitions of all the terms it uses. Here's the definition of "climate change" and correlated terms as used by them and, by extension, by climatologists, as of 2018's 5th synthesis report [www.ipcc.ch] (the glossary begins at page 133):
* Climate:
"Climate in a narrow sense is usually defined as the average weather, or more rigorously, as the statistical description in terms of the mean and variability of relevant quantities over a period of time r
Re: (Score:2)
You seem hysterical.
Hey dumb ass, what happened to whining about ad hominems?
You're about as sharp as a pet rock.
Um ... (Score:2)
> As for trees I can only assume you are joking. You..actually think there are more trees than there were 50 years ago? I don't know what to say about that.
So you'd fairly certain there are a lot less trees, particularly tue big ones, like rainforest? Correct?
Re: (Score:3)
Uhm, hate to rain on the parade here but a quick googling actually says we have more trees today than we had 35 years ago. Don't know about fifty but I think you may have pulled that number out of your beehind so I'm not sure it matters.
While what I've found also states that a lot of forest has been replaced by manmade plantations which don't have the same bioiversity, I'm not sure that factors into the discussion you're having...
Re:Chaotic systems and humility (Score:5, Informative)
I did pull it out of my ass. But the fact is we have, by one study [chicagotribune.com] 46% fewer trees than before humans started deforestation. Actual amounts will vary decade by decade but the fact is that humans are now the main driver of how many trees there are. We could deplete them to a tiny sliver, or we could assist or at the worst not interfere and they would grow back
We are pulling carbon out of the earth and spitting it into the atmosphere, and cutting down, burning, and replanting the same trees over and over doesn't really change the calculus.
Re: (Score:3)
Uhm, hate to rain on the parade here but a quick googling actually says we have more trees today than we had 35 years ago.
It may be true, but sadly it's irrelevant. Mature trees fix carbon faster than young ones because all growth occurs in a thin layer beneath the bark called the cambium, and because the rate of respiration is related to leaf area. The number of trees means nothing in and of itself.
Re: Chaotic systems and humility (Score:3)
Re: (Score:2)
A lot of Western movies were shot in my area years ago, when the town was mostly logged-off ranchland used for grazing. Today when we compare given landscapes with the old movie versions, the big change is that the trees have grown back.
Re:Chaotic systems and humility (Score:4, Informative)
It's a different problem. Modelling climate is a lot easier than modelling weather, because you're only interested in average, not details.
Re: (Score:2)
Re: (Score:2)
This. What's happening is irrelevant -- it's the holy grail for command and control, as class warfare dies away as successful reasoning. This was first noticed in the late '60s. The correct answer is society will compensate for it rather than ignore it or issue grotesque overpowers to government.
Re: (Score:2)
We can't even predict the next coin toss. Let's cancel statistics.
Re: (Score:2)
Sure we can. 50% will be heads, 50% will be tails.
Not analogous with the Three Body Problem, or the climate.
Re: (Score:2)
Re: (Score:2)
Not analogous with the Three Body Problem
And this is where you're completely wrong. For example we can't predict the trajectories long-term in a general N-body system, but we still have the virial theorem. Which is the equivalent of "50% will be heads, 50% will be tails" for N-body systems.
Re: (Score:2)
You still can't predict if the next coin tossed will come up head or tails. I don't think you understand the problem.
Re: (Score:2)
The same way, crash tests are not here to prove accidents happen. They are there to check how exactly the cars bend, and how hard the steering wheel hits the head of the driver.
Re: (Score:2)
"Climate Change" as you are using it is unfalsifiable.
The climate constantly changes, by definition.
It is precisely the proposed consequences that make the term have any significance. Demonstrate specific consequences, and your cost-benefit analysis of that outcome for remediation strategies that can be -objectively- considered.
Until then, you're just handwaving with one hand and picking people's pockets with the other.
Re: (Score:2, Insightful)
Re: (Score:2)
You should look at what reinsurance companies do. They clearly see the economic damage and plan ahead for more. These companies have to pay for some of the economic damage caused by climate change (but most of the damage will not be insured for). Munich Re - the largest reinsurance company worldwide - already warned about climate change in the 70s. Nowadays, many of the reinsurance companies divested from coal-related business and some even stopped to insure coal plants.
Re: (Score:2)
So, economic damage paid for by the people who get paid to take on risk of damage.
Works for me. Still not seeing why I should pay for some other country's unearthed corpses or obligations I never agreed to.
All this is besides the point, though. There will be no attempt to be "fair" in any sense like this. It will be a pure power and money grab, exactly as Communist International put on their short list for things to try once "we must enslave the worker for the benefit of the worker" was running out of co
Re: (Score:2)
Yes, I live in Austria.
Re: Chaotic systems and humility (Score:1)
Re: (Score:2)
We can barely calculate the future motion of three perfect spheres.
Yet, the future interaction of hundreds of "Climate Change" causal factors is "settled science".
You only have 10 fingers, but can type 36 letters and numbers.
What's your point?
Re: (Score:2, Insightful)
We can calculate it just fine. It just takes a lot of computing power.
If you insist on drawing an analogy to climate science, choose 3 arbitrary planets in free space with no outside influence. The climate change denier is the one shrieking that the planets won't move at all because gravity is just a theory while he plummets to the surface of the nearest planet.
Re: (Score:2)
Then of course, comes the language lawyering even though we all know damned well what was meant and what was said.
Re: (Score:2)
We can barely calculate the future motion of three perfect spheres.
Yet, the future interaction of hundreds of "Climate Change" causal factors is "settled science".
Nope. And nobody sane is claiming that. (No, _you_ are not sane.) This is risk management, and the little problem cretins like you gloss over is that the current models could also be optimistic.
Re: Chaotic systems and humility (Score:3)
You barely know the the chemical-electric impulses that move your legs, yet youâ(TM)re able to walk.
Still, the more you would be conscious about it the better youâ(TM)d be in moving yourself around - perhaps youâ(TM)d walk more elegantly and use slightly less work to do so.
The trend lines will not be inverted from the new climate data and information that are gathered.
There will not be an aha moment like hey wait a minute we got it all wrong.
And actually, we can already compare earlier models
Re: (Score:2)
We can barely calculate the future motion of three perfect spheres. Yet, the future interaction of hundreds of "Climate Change" causal factors is "settled science".
If you haven't yet encountered a system where the the microstate of several individual components of the system is more difficult to predict than the macrostate of the whole system, then I wish you the best of luck when you finally get into high school.
Re: (Score:2, Insightful)
We solve multi-body problems all the time for space applications - just not analytically. We converge on practical numeric solutions for each case. Analytic solutions tend to be toy solutions for idealized problem descriptions, just like our early climate models. Improved climate models are being solved numerically, just like the path of a probe through Saturn's moons.
Re: (Score:3)
In one where the other bodies are either too far away or too small to influence the result by more than a fraction of a percent. So sun-earth-moon would work perfectly fine.