Comment by SamoyedFurFluff | Hacker News Reader

Comment by SamoyedFurFluff | Hacker News Reader

SamoyedFurFluff

Understand that the papers emphasize they are not predictions, but illustrating complex relationships. (Or at least trying to illustrate complex relationships based on what we know of them right now). There is very little we as an individual level can do about whether or not full scale technological advancement occurs enough to reach steady state of equilibrium on a global level. It’s very important cognitively to realistically assess your personal ability to influence or not influence a global phenomenon.

Personally? I want to make sure my local community is strong. Even if no collapse occurs, a strong community is good. But if anything happens, a strong community is great. I do food coops (I buy a share of a farms crops ahead of the growing season, so the farmer shares the harvest with me. Sometimes I also pitch in to pack the shares for others in the coop!). I donate to community food availability(not just food kitchens but also paying way more for pay-what-you-can produce). I attend community events like cleanups, town halls, etc.

Most of us in tech are lucky enough to be able to work from the neighborhoods where we live, and be able to step out for the occasional chore. We inherently have more time to directly benefit our neediest. Why not take advantage?

> the papers emphasize they are not predictions, but illustrating complex relationships

If it has no predictive power, why should it be taken seriously? I'm concerned about the environment and sustainability too, but that's the exact reason we can't afford to throw empiricism out the window the second things start looking a bit hairy.

A paper not containing predictions is not the same thing as its having no predictive power!

This paper makes a lot of predictions despite the lack of predictive power. Hence, it should not be taken seriously.

Since the 1970's the group producing it has made a lot of very wrong predictions out of the hubristic belief that they can reduce everything about life to a simple system of differential equations.

What? Maybe my brain isn't big enough to understand this, but I don't see how you could disentangle the two without getting into pedantry.

Paper may say you can use this methodology with data beyond the scope of the paper to make a prediction.

However it doesn’t actually do the work to predict anything.

Kind of like the difference between theoretical math and applied math.

Take a look, for example, at the timeline in https://xkcd.com/1732/. It contains (almost) no prediction, but once you have internalized the information in it your own predictions on the climate topic are likely to become a lot more powerful because better informed.

Unfortunately for all of us, the model does have very good predictive power despite its simplicity. Herrington et al in 2020 checked the 1972 model scenarios (there were 4 such scenarios as sensitivity analyses) against around 40 years of empirical data and the results were close to observed ground truth [0].

0. https://onlinelibrary.wiley.com/action/downloadSupplement?do...

I try not to be too harsh on here, but your interpretation of this analysis is unbelievably bad. It's absurdly easy to fit a logistic function to match the real-life logistic curve of some value of interest during the exponential phase of those functions.

However... That means almost *NOTHING* with regard to the actual logistic portion which is the part people actually care about. There exist an infinite range of different logistic functions that can be made to match in the exponential portion and the majority of those functions contain almost zero information about the future logistic portion. The error between them is essentially unbounded.

Calling that confirmation of predictive power is not misguided, it's just about as wrong as you can be.

This is not my analysis, and the first author, Gaya Herrington, explains the alignment between the model and empirical data here: https://m.youtube.com/watch?v=KJpvc3i5j84

I also react with concern. Yes there are infinite curves, but there are limits to what is possible, so in actual fact there are not infinitely many possible futures here, and those futures are likely to fall within the limits of the 4 sensitivity analyses…and indeed have over more than 4 decades of follow up.

I know it's not your analysis. I said your interpretation of this analysis is flawed. You're misunderstanding that analysis to be confirmation of the predictive power of the model extrapolated into the logistic region. And you still are:

>Yes there are infinite curves, but there are limits to what is possible, so in actual fact there are not infinitely many possible futures here, and those futures are likely to fall within the limits of the 4 sensitivity analyses…and indeed have over more than 4 decades of follow up.

The key phrase there where your interpretation is flawed is: "are likely to fall within the limits of the 4 sensitivity analyses". That's so, so untrue. You simply cannot extrapolate logistic model parameters into the actual post-exponential-growth, logistic portion without seeing the actual logistic effects start to kick in. Those simply haven't been empirically verified here.

And speaking of... the original model they are analyzing introduces limits on that exponential growth in a pretty arbitrary and artificial way, particularly when it comes to the pollution terms.... which, I'll add, this analysis shows is also the parameter they most over-estimated by a large amount even in the exponential portion.

I want to believe in infinite growth too. Maybe asteroid mining could make this paper and model predictions way off. Indeed, that is the kind of thing we hear from Jeff Bezos: https://www.milkenreview.org/articles/mining-in-space-is-com...

But also, yes, it IS reasonable to look at 4 sensitivity analyses comprising the most common rebuttals to limits of growth, observe alignment with more than 4 decades of empirical data, and believe based on that the model is pitching well above its weight. No need for ad hominem here.

You don't have to believe in infinite growth, the problem with the model that makes the scary bit after the slopes change to down are all BS because of: "However, it is important to note that the connections in the model and the recalibration are only valid for the rising edge, as many of the variables and equations represented in the model are not physical but socio-economic. It is to be expected that the complex socio-economic relationships will be rearranged and reconnected in the event of a collapse. World3 holds the relationships between variables constant. Therefore it is not useful to draw further conclusions from the trajectory after the tipping points"

That tells you this model disregards how every single societal crisis in the past was solved or at least mitigated: A massive change to the order of society (think germanic tribes conquering rome, magna carta type stuff, Napoleon et al taking France from a monarchy to an anarchy to an empire and, after calming down a bit, to a republic, etc. etc.).

Which of these graphs do you think supports the claim that LFG has very good predictive power? I browsed the first few, population, fertility, mortality, and food per capita. In each case the empirical data clearly diverge from all four LFG predictions, in exactly the direction I would have guessed. Or am I wrong in that analysis?

You are engaging in self-delusion here.

The point of these papers isn't to make precise predictions of "the day civilizational collapse occurs" nor does it merely "illustrate some complex relationships".

It is to show every closed system, even only remotely resembling our own, is bound to end in overshoot&collapse unless you change central tenets of the driving economic paradigm. Namely abandon exponential growth for a steady (or oscillatory) state approach.

Individuals aren't helpless bystanders either. As adults in a democracy, it isn't only your right but your responsibility to bring about the necessary change.

We don’t live in a closed system though.

Looks pretty closed to me: https://earthobservatory.nasa.gov/images/144427/all-of-you-o...

How many thousands of dollars does it cost to launch a kilogram of mass into orbit again? Out of orbit?

Do we even have a closed system that can support a small group of people indefinitely to launch? Unfortunately, we do not yet have such a system outside of the earth itself.

Earth's biome gets most of its energy from the sun. I think that's what the GP was getting at.

The physics definition of closed system allows exchange of energy. If neither matter nor energy can be exchanged, that's called an isolated system.

Thanks for the clarification. I'm not sure in what sense the others are using the term. The mechanism of collapse in the paper, exhaustion of fossil fuels, relies on not enough energy in the system, so I assumed energy was part of the calculation.

Since we aren't able to export our pollution out into space, or mine food on mars, yes, we live in a closed system. We live on a round spaceship known as Spaceship Earth (https://en.wikipedia.org/wiki/Spaceship_Earth)

Strictly speaking, even if we were able to do that we'd still live in a closed system, just a much, much larger one. :)

Yes, we do.

For all practical purposes, earth's gravity well serves as a boundary for material resources at scale.

More importantly, the input of polluting material into earth's obviously very limited ecological system is limited. You simply cannot pollute as much as you want, so getting more stuff from the asteroid belt or whatever doesn't help.

The whole point of voting in my country is to serve as a check on the legislature against using the government to do things to its citizens. The ballot box isn’t a tool for societal change.

A dictatorship is a much better system for changing society, if that’s your goal for government.

> The ballot box isn’t a tool for societal change.

Then what is?

I mean, to a degree I agree with the GP. You can't fundamentally change a system from within the system itself. No democracy has ever turned via a series of elections into, say, a monarchy. The only way that happens is either by violent collapse (e.g. a coup or a revolution) or by gradual erosion of bureaucracy (where increasing corruption enables a few powerful individuals to seize control regardless of what the law says). The point of elections are not to make dramatic changes, but to make tweaks.

> The point of elections are not to make dramatic changes, but to make tweaks.

The point of bureaucracy is to limit change, to not allow dramatic change.

The point of democracy is to make decisions. And the larger the groups of people and the less frequently voting happens, the larger the granularity of those decisions.

Decisions should be able to change bureaucracy, add to it or subtract or realign or whatever.

When bureaucracy is used to gate-keep decisions that can be made in a democracy, it's not really democracy any more. And I think a lot of people, regardless of their political beliefs, believe that has happened: nominal democracy where votes don't really matter because they can't affect the bureaucracy.

I would say the point of bureaucracy is to make the functioning of a system consistent with some set of rules. The preservation of those rules in some specific form does not have to be part of the rules, and so does not need to be a goal of a given bureaucracy. However, a self-preserving bureaucracy can arise, either by design or accident. I would say every government that has ever existed fits that description. Can you think of a single one that has not been fundamentally conservative? "Conservative" meaning seeking to maintain the status quo that allows it to exist in its present form at any given time.

You didn’t address that I already participate in democracy. I’m just realistic in my actual capacity to execute change, which is inherently largely local, so I focus on local change.

That is assuming exponential growth inevitably leads to overshoot & collapse, which is a flawed zero-sum perspective. Growth (adding value) doesn't necessarily have to come at the cost of resources.

Reply on news.ycombinator.com