Hacker News Reader

Here are two scary storylines that get discussed pretty regularly here on HN:

a) AI and machine learning will raise productivity so much that they will steal millions of jobs from people.

b) Global population will eventually go into decline, which will crash the economy by killing growth.

But, these two stories look far less scary when considered together. In fact they seem directly complementary, although the timing is not likely to work out perfectly, so it won't necessarily be smooth sailing along the way.

now if we could just get the robots to pay social security

You jest, but the economic effects of all of this are severe. A total regression in the real-estate market (because stagnant demographics means the existing housing stock will roughly suffice or even exceed demand) will have severe effects.

Basically the end to growing demand, and thus essentially an end to growth-orientated capitalism.

I only half jest, but this demographic situation is one of the reasons I support liberal immigration policies.

That fails when every country has falling birth rates

Swedish authorities did a study about this for that very reason. The conclusion was that it's unlikely that people will want to move halfway across the World only go get a job changing old peoples' diapers.

Alas, the PDF seems to have been taken off-line.

My long term prediction is that once the boomers start dying off in droves, we'll face a real estate crisis as their houses flood the market...and many less desirable neighborhoods throughout suburbia will turn into mini ghost towns.

I come from northern Italy and during my relatively brief lifetime (length(life.qubex)>40) I’ve seen this happen to formerly exclusive “Beverly Hills”-type affluent residential areas (and un/gated communities). Enormous, cavernous hangar-sized homes that used to be worth millions of euros now struggling to sell for a couple hundred thousand.

You're describing Japan, now. Large numbers of abandoned houses, towns in rural areas emptying out, and the process is accelerating. An interesting article: https://spikejapan.wordpress.com/2014/08/16/yubari-withering...

It’s almost as if we should redistribute profits from the robots universally to provide some level of basic income.

It's already happening though - commodity/consumable prices are far lower in real terms today compared to the previous decades.

Firstly that has nothing to do with redistribution.

Secondly total cost of living in a major city has gone up, not down, due to rent and realestate. Sure, i can but more polypropyl, bread and stainless steel, but what good is that if i have nowhere to live.

> now if we could just get the robots to pay social security

It's called “taxing capital” and we can (logically; whether we have the political will is a different question.)

Political will quickly arises if there is loud public demand for it.

That is not true in the US. What happens in the senate has low overlap with what the public wants compared with what donors/lobbyists want.

US economic policy has an almost null correlation with public demand (and interest, really.)

Well we can. Someone owns the robots (or owns shares in the corporation that owns the robots). Just change the tax code to make Social Security tax apply to all income, including capital gains, with no upper limit.

The reason why capital income is not subject to social isurance taxes is because it is fundamentally an insurance scheme. People who earn mostly wages need social insurance in retirement (when they would not be physically able to work), so they (have to) pay it, while people who earn mostly capital income do not need it (as capital income does not depend on personal health / fitness to work).

Therefore extending social insurance taxes to capital income without extending pension payments to capital earners would be unfair. Local politicians tried something similar - remove caps on social insurance, while keeping pensions capped / strongly sublinear, and it was struck down by constitutional court.

Perhaps reasonable solution would be to split pensions to tax-based UBI, and smaller contribution based 'linear' pensions.

C) We will develop anti-aging treatments due to increased funding of SENS and similar projects. Combine that with a universal basic income and robots and you get a much better picture of the future.

Edit: we are also not taking into account artificial gestation methods that may be developed in the next 100 years. In 2120 it may seem as barbaric as child labor and other 19th century atrocities to actually have a woman carry through a pregnancy, not to say that religious fundamentalists will want to continue the practice.

If we are able to keep people alive for much, much longer but don't find a way to make their brains young and elastic again, we're going to be in for a very, very special kind of dystopia.

“Anti-aging” implies keeping people younger, not just delaying death.

If AI takes jobs, it doesn't cause wealth to be redistributed, it causes wealth to flow upwards to the person who owns the AI.

That might then be redistributed, but it might also not. Whether it is or isn't is largely up to the people who have the most power. In that system, which group has the most power?

But if AI is intelligent, is owning AI slavery?

If an AI is self-improving, why should it have an owner?

If AI improved itself, who owns the copyright?

"Should" doesn't matter. All that matters is how power is actually distributed.

Thats what the French aristocracy said, and eventually they found their head on a pike.

Sure, because the populace had more power than they thought, and they organised. It didn't matter how rich the aristocracy were, they couldn't buy enough loyalty to defend themselves. Down the line you probably will be able to buy a robot army, and robots aren't squishy.

In democratic societies the majority decides how resources are distributed. If they can understand where their interests actually lie it shouldn’t be a problem.

If you put a referendum to the people tomorrow, saying "should we tax billionaires more than other people, and spend it on you," you'd get an overwhelming yes. The fact we don't have policies like that show the majority aren't deciding that distribution.

But you're also assuming society stays democratic. Why would it, if enough power is aggregated at the top? Who's going to stop them if they decide to form a cabal?

Or consider the alternative scenario where population keeps declining but technological advancement doesn't materialize. Which for AI and machine learning could happen soon with another 'AI winter' after people realize only narrow set of opportunities to deploy it.

The stability of society depends on technological progress and more children born in each new generation. Once this global Ponzi scheme reverses there is an incentive to manufacture lies about massive technological progress ahead of us. See buzzwords such as autonomous cars, industry 4.0 or quantum computing.

Even with another AI winter I can’t see that automation isn’t going to continue for a long time.

AI is just one piece of increased productivity. Even without AGI there are paths that could lead to 3x productivity faster than the population would go down by 66%.

There are other issues with an older age distribution which cannot be fixed by productivity though. There could be an oversupply of 5 bedroom homes for example.

Counterpoint: major world problems like climate change are a direct result of having too many people, and more specifically having too many people living unsustainable lifestyles. Worse, a large percentage of the world's population is now moving out of sustainable living into unsustainable living as a result of globalization (as people gain more wealth, they tend to want things like more meat and driving a car).

The net result is that a decreasing population is a good thing for the planet, but a bad thing for economics.

> Counterpoint: major world problems like climate change are a direct result of having too many people

Not exactly. The bottom 85% (in income terms) of the global population contribute very little to unsustainability. There are too many people in OECD countries, perhaps.

> The bottom 85% (in income terms) of the global population contribute very little to unsustainability.

Is that true? What percent do they contribute?

This should give you a rough idea.

https://www.overshootday.org/newsroom/country-overshoot-days...

> Not all countries will have an overshoot day. By way of the country overshoot equation above, a country will only have an overshoot day if their Ecological Footprint per person is greater than global biocapacity per person (1.63 gha).

There’s a lot of countries not listed, i.e they do not overshoot.

I'm still wrapping my head around the meaning of those overshoot days.

A person's ecological footprint is a multifaceted thing for which I think a single number cannot necessarily provide a very meaningful summary.

But here's some other data, that considers only CO2 emissions: https://www.ucsusa.org/resources/each-countrys-share-co2-emi...

It groups by country rather than income level but, reading between the lines, it seems to me almost certain that the bottom 85% (in income) of people emit more than half of the global total.

BTW it is important to consider that a large fraction of the emissions from countries like China and India are in the service of consumption in wealthy western countries. But, I googled, and even in a famously export-heavy country like China, exports only account for 17% of the GDP.

That is also partly because bottom 50% don't have decent lifestyles. Try commuting for 1.5 hrs in packed local trains in Mumbai and you will definitely wish for a slightly higher carbon footprint in exchange for a better quality of life. Off course, USA standards is unsustainable for the world, but at the least Japanese standards of living are like human rights.

The bottom 85% of the global population contribute very little to unsustainability, TODAY.

They are going to dominate unsustainability in the next few decades.

More coal fired power plants were built in Africa in the last decade than all other continents combined.

The rapid and massive increase in the number of global middle class is why virtual everything will be important for the planet. If virtual services and experiences become almost as useful and enjoyable as the real world, people will burn less fuel and cause less negative externalities.

(They could be better as well! Even those who enjoy traveling, for example, don’t care for the hassles of airport queuing, getting stuck in sardine-like plane seats, finding laundry places, etc.)

Obviously, real-world experiences necessary for genuine, long-term human connections should remain.

Note: Although the birth rate is falling, the number of global middle class is expected to rise a great deal in the next few decades, largely due to rapid economic development.

This is very interesting, and unexpected at the surface level. Especially with all the doom and gloom we've always been fed about the world population getting out of control, consuming all the resources, polluting the planet into oblivion, etc.

I was expecting to read about medical/health reasons why men and women are literally becoming less fertile, but it's not that. It sounds like it's a side effect of countries/societies advancing technologically and socially. It makes sense, but I just never visualized it like that in my head. Wow, what a great problem to have!

But as you say, it is a problem, just in other ways we might not have considered before. It will be interesting to see how we figure out how to restructure society as the demographics shift. But this seems like it could be a very good thing for the environment, climate change, resource management, etc. Overall, seems like positive new to me!

> Especially with all the doom and gloom we've always been fed

Which is still true, if the billion Indians and Billion Chinese start having American lifestyles, that is just not sustainable for the planet.

I fully expect the Indian and Chinese population coming into wealth will go through a period of fairly outrageous conspicuous consumption. I have a hope that it will be fast-forwarded compared to the past 70 years of American history. Hopefully after a decade-long warmup and another decade of settling in, the novelty will wear off and there will be a move away from focusing on "More Stuff!" and toward focusing on "Less Clutter, More Experiences!"

If those experiences involve moving mass long distances (such as flying to South America and hiking the Andes), then it’s all the same.

I don’t see any environmental reprieve short of drastic reductions in consumptions, including travel.

Good thing that high-speed rail and other land transport infrastructure is being built in order to connect the entirety of Eurasia and eventually Africa, with sufficient throughput to feed multiple US economies.

> Hopefully after a decade-long warmup and another decade of settling in, the novelty will wear off and there will be a move away from focusing on "More Stuff!" and toward focusing on "Less Clutter, More Experiences!"

The western world hasn't even moved past that stage, so I highly doubt developing countries will move through it quickly.

Ironically, it is cheaper to have "more stuff" than to have "less clutter, more experiences!".

India and China are both projected to have a <1B population by 2100.

I think by 2100 we have either solved this problem or is too late. I guess most likely the latter but there is still hope.

These projections do not account for increase in life expectancy. India is projected to peak at 1.7 b, if you consider Indian subcontinent as an entity, it puts the population of South Asia at 2.25 b. Not a good sight when coupled with unreliable climate, water scarcity, unlivable summer temperatures and no direct opportunities which means corruption may lead to feudalism of the yore.

> These projections do not account for increase in life expectancy.

That's inaccurate. The study in the link, which has those predictions:

https://www.thelancet.com/journals/lancet/article/PIIS0140-6...

See the Mortality section.

Yes, India will peak at 1.7 billion, but then begin to decrease. That peak happens in 2050, which is closer to our 2020 than 2100.

Heh, they killed a lot of their women so that really puts a cap on how quickly they could grow.

I suppose that is a possibility. Declining population doesn’t necessarily guarantee declining resource usage.

  No one planned it or made it happen

are you serious ? UN planned this since the 70' there's policies and support for developing countries at the global scale for implementing birth control

Hmmm... Maybe.

It seems to me that birth control itself was developed for the needs of rich countries. And, the UN pushed it into poor countries because they obviously needed it too. That did help reduce unwanted pregnancies. But, until the reduction in child mortality and the mass move of couples to cities, families were still motivated to intentionally have lots of babies.

The "population implosion" problem is only so because of the economic obligations that will be passed on. I'm actually surprised that the US/UK elite, with their grand history in funding Eugenics research, didn't fund studies that'd portend this uncertain fate - then again may be it's just not published.

this is fantastic news. Yes, we don’t know how to properly adapt yet, but this planet resources are limited and we aren’t good at sharing them.

What worries me is that long term this can be reversed just like current trend is reversing away from population growth. Current trend seems to be driven by job prospects, education and costs of raising kids. But what if in a future we all had basic income? No need to worry about food, shelter or education for your children. Won’t this lead to population growing again?

Why would it? Do you imagine people enjoy raising four kids or more? Children used to be old-age insurance. You needed so many to be certain some of them would survive, thrive, and take care of you in old age. If your needs are met, either through a good pension or UBI, you can have as many kids as you prefer. And by all indications, for most people that's actually around the natural replacement level, or fewer.

> for most people that's actually around the natural replacement level, or fewer.

Not sure where that number comes from. 2.5 is probably closer to the natural trend. Physical constraints (food, disease, etc) limited human expansion previously. The economic constraints that cause a min-max of not having kids will never be A. universal, B. persistent. A couple generations of baby recession won't impact the fate of humanity one bit. Not sure why there's so much faux concern over a fixable (and likely temporary) situation. You give a tax break of 50% on capital gains per child and the kids will be popping from the higher economic rungs.

> You give a tax break of 50% on capital gains per child and the kids will be popping from the higher economic rungs

Of course, with a perversely positive incentive like that. But it tells us nothing of the natural impulse in a neutral environment.

In a lot of the western world there is plenty of support for having kids. Ample child care leave, solid monetary support, free or affordable child care and schooling. And in the countries which have most of this, the birth rate is low but not catastrophically so. (see Germany, the Nordics..)

Whereas in the US and Japan, for example, there are certainly structural factors that now weigh against having kids. But it does not tell us much about the "default state"

I don’t think there is such a thing as a default state. What would a neutral environment even look like?