Comment by ljw1001 | Hacker News Reader

Comment by ljw1001 | Hacker News Reader

To my mind, these are the most important, unsolved problems:

- environmental degradation is based on two exponential growth trends: human population and per-capita consumption (driven by economic growth). Solving either could be, literally, a save the world scenario. The first is easiest technically, (healthcare, non-coercive birth control, education, economic opportunity for women) but a political minefield. (yes, population is still growing exponentially)

- dealing with climate change. At a minimum, this means figuring out how to handle a significant rise in sea level in the next 100 years. Walls? Relocation? We need to think about response now, more so than avoidance.

- actually doing geo-engineering to prevent/reverse the greenhouse effect

- preventing a nuclear war

- preventing a large asteroid collision

- dealing with large-scale disruption of employment (for example retail continues to be hammered by online stores, self-driving trucks, etc.) Build a better Mechanical Turk might be a useful strategy.

Obviously some of these are low-probability short-term events, but they have very high costs. In a sense, you would be taking advantage of people's tendency to systematically under-invest in problems of that kind.

Fun fact: not only population is no longer growing exponentially, it doesn't even grow linearly. The growth is now _sublinear_, and it looks like we'll have peak population around 2070. Problem solved. Let's do something more interesting.

Frome wikipedia: "Global human population growth amounts to around 75 million annually, or 1.1% per year." 1.1% is exponential growth when the base is growing, just like 1.1% compound interest displays exponential growth. It is still growing exponentially, although the rate is declining.

If the rate is declining, the growth is no longer exponential :)

picture two exponential growth curves, one steeper than the other, over time the a trend moves from the steeper curve to the less steep curve. The "population" whatever it is, is still growing exponentially at the less steep curve.

The idea is that an exponential curve is no longer the correct model for forecasting growth past 2070.

Arguing about how to express that is sort of pointless.

Population growth in non-human species is quite well documented and is in fact not exponential, although the early stages look exponential.

It is in fact a logistic sigmoid function. It is amazing, in that it is not what you'd expect : clearly the vast majority of species have the ability to correctly estimate the carrying capacity of a habitat and only overshoot by a little bit.

So the Malthusian argument is incorrect, probably, if humans are like other animals (and since population growth curves in many localities do indeed look like S-curves I would say that this ought to be the default claim). Human population size will by itself stop just shy of the carrying capacity of the planet. Which is exactly what you'd want to happen.

In all likelihood, this will happen without active measures from governments (which won't work anyway due to the global nature of the required policies), with or without popularizing contraceptives, with or without female labor participation, ... (not saying those aren't desirable, just that they aren't necessary)

The alternative requires an assumption that humans are hugely different from the species that we evolved from, and since this certainly isn't true in the biological sense, why would our population growth be so very different ? Humans are not the first large animal to "conquer the planet", and reach population numbers in the billions, with some fish well on the way to trillions even today, and we aren't the biggest species in terms of total biomass by a very long way (amongst non-plants that honour likely goes to ants, but even pathetic species of bacteria dwarf their numbers). Will we be the last one ? Perhaps, but it seems unlikely.

https://en.wikipedia.org/wiki/Logistic_function#In_ecology:_...

It's not that species have the ability to correctly estimate the carrying capacity of the environment. It's that when the population of a species exceeds the carrying capacity of the environment, individuals will die and/or fail to reproduce, dropping the population of the next generation.

It's very likely that humans exhibit the same behavior, but on a micro-scale, this isn't very comforting. If a politician says "Well, the earth will be fine, but 10% of you are gonna die and another 30% won't be able to have children that survive to adulthood", there would be rioting on the streets ... which has a non-negligible chance of bringing about those numbers.

Nature will be just fine, but remember that the mechanisms by which nature self-adapts are the four horsemen of the apocalypse: pestilence, war, famine, and death.

> So the Malthusian argument is incorrect, probably, if humans are like other animals

> The alternative requires an assumption that humans are hugely different from the species that we evolved from

Well we are, for proof look at the fact that we're even discussing it here when all the other species never did. Look, that all sounds very peachy if it weren't for the fact that our species alone has grown the brains and the self-justifications for consciously attempting to continously outwit "nature" (and be able to replicate way beyond what our ancestral habitats could carry) at every turn with agriculture and all the sciences =) granted, there's no point in or way of "going back" for all manner of reasons, but we needn't lose sight of certain basic facts here..

I refer you to the 12,000-year human population growth chart on Our World In Data, in a post by Max Roser.

https://ourworldindata.org/world-population-growth/

Growth is clearly exponential. Your theories don't fit the data.

I'm looking at that graph. I see a spike in the growth rate, after the spike it goes to zero. I'm not an expert (I have a masters in statistics actually) but that doesn't look very exponential to me.

Can you explain what I'm missing ?

2% is exponential growth; 1% is still exponential growth. Decrease from 2% to 1% is a decrease in the rate of growth, which still remains exponential.

(Sigh).

Look at the linear function. It experiences "decrease in the rate of growth" along the line.

If the exponent is not constant, we do not have "exponential growth" anymore — by definition.

picture two exponential growth curves, one steeper than the other, over time the a trend moves from the steeper curve to the less steep curve. The "population" whatever it is, is still growing exponentially at the less steep curve.

LOL. somebody downvoted this, but you know. Math.

People like to use percentage for everything. That doesn't mean that everything is exponential.

No, but we are talking about a formula of the type where a growth rate G is a factor of a base population P, and P in Period n+1 = P(n) + G*P(n)

In this case G is the difference between the birth rate and the death rate for time period t. As long as a G is positive, you get exponential growth. Every time. This is how compound interest works. The fact that your balance next year is based on your starting balance + the balance multiplied by a positive growth rate ALWAYS leads to exponential growth in the balance.

Here's a video that explains it. https://www.youtube.com/watch?v=NB4KVSjgi24

> In this case G is the difference between the birth rate and the death rate for time period t

OK then. Let the birth rate be 1 person per second and death rate 0 (nobody dies). Do you get exponential growth? No. This is still linear growth.

Your another mistake is that ax^b is exponential growth, and ax^c is also exponential growth (if b and c are constants), but these are different processes. If you get a single process with changing exponent, this is no longer exponential growth, it can be arbitrarily defined.

Thank you. You finally made it clear exactly what you don't understand.

One person per second is linear.

.00001 person per year * current population is exponential, assuming only that the (fractional) person is added to the population factor so that the base is bigger next year.

If you don't understand that, you shouldn't be commenting on growth rates.

Indeed. However you also made it clear what you don't understand. Consider 1 person per year being born.

If you assume the growth rate is exponential, then you will see that the first year, the growth rate is 100%, the second year the growth rate is 50%, the third year the growth rate is 33% and so on. At any point, you can fit an exponential to the graph tangential to the current point. It won't fit much, but you can still do it.

A decreasing exponential may be exponential, but it also very well may not be. But we'll need to look a few years into the future to be sure. That said, expert opinion seems to be that growth rate is at this point linear (or iow an exponential that decreases each year so as to appear linear) [0]

[0]: https://ourworldindata.org/wp-content/uploads/2013/05/update...

> A decreasing exponential may be exponential

No, a "decreasing exponential" is not exponential. It might return to an exponential after the decreasing period, but during the period of decline, the growth is not exponential.

Fair.

The population growth rate isn't dependant on the base population. You can pretend it is by expressing it as a percentage, but that doesn't make it so.

Look at [0]. 2016 had a higher population than 2015, yet the growth was smaller. Not just a smaller percentage, but actually smaller by about half a million people. If growth was exponential, then the higher population would have a larger increase, but it doesn't, and so the growth isn't exponential.

[0]: http://www.worldometers.info/world-population/world-populati...

Sigh. The growth _rate_ is not dependent on the population, but the actual numeric growth _amount_ is. It is the _population_ that is growing exponentially.

You are describing the shift from a steep exponential growth curve to a lower exponential growth curve that I mentioned earlier.

If you get 10% compounded interest on your money it grows exponentially at 10%.

If you get 9% compounded interest it grows exponentially at 9%.

And if you _were_ getting 10% and the bank changes the rate to 9% your money is now growing exponentially at 9% interest, assuming only that the interest is reinvested.

Assume my bank gives me a linear interest of $10 per year.

Then going from $100 to $110 that would be 10%.

Going from $110 to $120 would be ~9%.

Going from $120 to $130 would be ~8%.

In any year you could conclude my savings grow exponentially. 10%! 9%! whoohoo! But if you look over a longer time frame, it clearly isn't.

And as nevdka mentioned, the population growth is less than linear. If my bank gave me $10 of interest last year and $9 this year, and $8 the next, there is definitely no exponential growth.

Well, if over time it is continuing shifting to match slower and slower exponential curves, its not actually exponential growth over the long-term.

A growth curve that is, in fact, logistic, would look like that.

I never said that it would continue to decline indefinitely. I said it was exponential before, and it's still exponential, even though the rate has changed. In spite of the massive negative response here, that statement is perfectly true.

> I never said that it would continue to decline indefinitely.

It's been continuously declining for half a century after peaking, the assumption that it's suddenly reached a bottom at the most recent measurement and returned to an exponential steady state is unwarranted.

"and it looks like we'll have peak population around 2070"

So how do you factored in all the possible coming or not coming atomic/biological/chemical wars + new breakthroughs in medicine/agriculture?

And since right now "More than 100 million people worldwide are facing acute malnutrition and risk starving to death" [0] I would call it a bit cynic to say "Problem solved".

[0] http://www.independent.co.uk/news/world/politics/number-of-p...

The answer to solving world hunger is not necessarily reducing human population. Are you sure it is me who is cynical here?

Additionally, more humans are expected to die from obesity than from hunger, as of now.

"The answer to solving world hunger is not necessarily reducing human population"

I did not say that. I just said that the problem is clearly not solved, when so many people are starving.

The items you mention are intellectually interesting, but somewhat counterintuitively, do not influence population growth in the way that one might expect. Case in point: breakthroughs in medicine and agriculture.

One might expect that access to food and medicine leads to population growth, but in the long term, countries with the highest levels of access have the lowest population growth. It's quite a strange phenomenon, and it is basically this trend which leads statisticians to estimate 2070 as having a peak population.

"It's quite a strange phenomenon, and it is basically this trend which leads statisticians to estimate 2070 as having a peak population."

Yes, but it might be a missleading one. Because I think our society causes many people to consider children as expensive and time-consuming and therefore not wanting to have children.

But that might change again. There were times when childrens were be seen as a luxery or as a necessity, a insurance for old age.

I mean, I am not saying it will change to that or that direction, but I think you can't really predict the future on how the world is acting today as human society changed quite drastic in the last hundreds years. I would rather predict that human society will continue to change and therefore I am very cautious with any fixed predictions.

Children are expensive and time-consuming, because in our society we're expected to take care of them, and the standard for that is very high. In developing countries, the standard is much lower and children are able to help run the household, so they're incentivized to have more children. As the things that make a country 'developed' spread in developing countries, the equation starts to tip the other way...

This is definitely the line of rationalisation I'd use to explain the trend. The person replying to me wasn't entirely wrong in questioning whether it would continue, however.

E.g. I can certainly envision post-work economies, where the "penalty" of having multiple children is much lower, as they are essentially guaranteed a good life. I wonder which direction society's value system would evolve in that scenario.

Are you talking about the west or the whole world?

Because if you're talking about the whole world, I'm pretty sure you're incorrect.

The trend is universal, although some developing countries have not reached that stage yet. Kurzgesagt has a good video on this: https://www.youtube.com/watch?v=QsBT5EQt348

The trend is towards low/no growth, but it's not there yet because of the current growth in some developing countries as you mention.

People need to recognize that this is, currently, just a trend, and not yet a fact and certainly not a physical law. For example, large families are correlated with religious fundamentalism, which is rising. Could it rise to the point where the trend changes? It is possible.

Family size is generally going down everywhere, also in religious countries. Population growth is mainly because we live longer.

It's not a physical law, but still pretty well understood phenomenon: more likely your offsprings are going to survive, less likely you are going breed more of them.

Rosling also claimed that population growth projections have been very accurate for decades.

As "we" have fewer children and "they" continue to have more children, I'm not sure where the future is headed. I'm not saying Idiocracy levels but you have to wonder if the video is correct.

For example, what is erdogan thinking when he says Turks should have five or more children? I'd love to be able to read what's going on in people's heads. Personally, I couldn't give two (insert expletive) if my "culture and heritage" were to disappear or dissolve I to something new in fifty or a hundred years.

Global growth rate is dropping... whether it's sublinear depends on where you want to start the clock and what resolution you want to consider, I think.

http://www.worldometers.info/world-population/#growthrate

The rate is dropping but the population is still growing and will continue to do so for some time.

Wholeheartly agree. Let me extend your list.

In 2004, the United Nations' "High-level Panel on Threats, Challenges and Change" analysed threats and challenges to international peace and security. Their top ten list in priority order: Poverty, Infectious disease, Environmental degradation, Inter-state war, Civil war, Genocide, Other atrocities (e.g., trade in women and children for sexual slavery, or kidnapping for body parts), Weapons of mass destruction (nuclear proliferation, chemical weapon proliferation, biological weapon proliferation), Terrorism, and Transnational organized crime. [1]

Jean-francois Rischard's "High Noon: 20 Global Problems, 20 Years To Solve Them" also identifies 20 issues broken down into 3 groups [2]:

Group one (sharing our planet) includes global warming; biodiversity and ecosystem losses, fisheries depletion, deforestation, water deficits, and maritime safety and pollution.

Group two (sharing our humanity) includes massive step-up in the fight against poverty, peacekeeping-conflict prevention-combatting terrorism, education for all, global infectuous diseases, digital divide, and natural disaster prevention and mitigation.

Group three (sharing our rule book) includes reinventing taxation for the 21st century, biotechnology rules, global financial architecture, illegal drugs, trade-investment-competition rules, intellectual property rights, e-commerce rules, and international labor and migration rules.

edit: formatting.

===

[1] https://en.wikipedia.org/wiki/High-level_Panel_on_Threats,_C...

[2] https://www.amazon.com/gp/customer-reviews/R28KHPSYSQ7O6T/re...

"- environmental degradation is based on two exponential growth trends: human population and per-capita consumption (driven by economic growth)."

You left out one detail: environmental degration per consumption. This is a really important factor, and is probably easist of th tree to minimize, and is being minimized every year. (Assuming that you didn't equate consumption and environmental degration).

yeah, it's not a complete model. :)

If you first instinct is to address these problems with some sort of startup, I think you've got a hammer-problem going on. These are societal problems that the market has shown no appetite for tackling or capacity for actually resolving.

The original post was about frighteningly difficult problems. I think they meet that criterion.

> If you first instinct is to address these problems with some sort of startup, I think you've got a hammer-problem going on. These are societal problems that the market has shown no appetite for tackling or capacity for actually resolving.

I'm cautious with this: If you know what problems people/society has, you know what products to solve/mitigate these problems might be bought. This perfectly sound like startup opportunities to me.

Unfortunately, all but the last item on your list have no or negative revenue potential. No one (save perhaps a few lefty billionaires) will pay you for the environmental stuff you mentioned, preventing a nuclear war, or preventing a large asteroid collision. I suppose there might be a small NASA contract in there for the asteroid issue, but the current budget for even tracking asteroids, much less stopping them, is small.

No, the biggest problem is how you monetize that. We already know how to solve these, but the short term cost/benefit for the averge person just isn't there. Figure out how you get average people to willingly spend their money for products that solve those problems.

add preventing epidemic that can wipe out humanity

Reply on news.ycombinator.com