it’s getting really hard to do macro analysis these days. either i’m bad at it, or things are really changing. i think my subjective probabilities of myself being bad at it is around 40%.

so the other possibility is drastic changes incoming. i mean that’s not even controversial now. back 10+ years ago when i intuited that capitalism-as-we-knew-it was going to collapse, AI passing Turing Tests wasn’t even on my bingo list. (and AI really even isn’t the main reason why I think things are going to change, even though it makes things more obvious to the average lay person)

if you squint hard enough, you can see why we are almost at the threshold of abundance (or post-scarcity). starvation is mainly an issue of inequality in first world economies, not an issue of production. in fact, speaking of production, the main crisis of today is apparently that the world has too much production in the name of China, and the system cannot deal with the excess; and production is not the only thing in excess. everything is in excess. money is in excess. information is in excess. knowledge is in excess. intelligence is in excess. compute… well, that’s soon.

i’m not an economist, although sometimes sound like one when wearing my prophet hat (which i try not to these days). the holy grail of long term “investment” these days is to find something that’s actually going to be both in demand and supply constrained… and the answer seems to be… nothing? The thing that currently matches that description is computer chips used for AI, but if we talk long (5+ years) term that’s not going to last unless some weird artificial scarcity happens. The widely-anticipated and touted “scarcity” of AI tokens failed to materialize because of Chinese open models commoditizing it (without which the duopoly of the two major players in the US might have been able to corner the market and reap huge margins). You have some interesting moves in commodities markets (eg. copper) that might signal a short/medium term shortage, but those are things that will sort itself out in the long term due to market forces.

And things are still getting cheaper. And the effect of AI making human-equivalent labor much cheaper hasn’t really been reflected in real world statistics yet. Those effects are mostly projected. Of course some businesses are already devastated (eg. those doing general translation by employing people from countries with cheap labor), but AI’s real economic impact is maybe still a couple years down the road, when eventually you’ll get the massive white collar unemployment spike (no unemployment, no cost reduction from AI). and given AI is still an empirical scientific experiment today, it’s hard to forecast how things are going to play out exactly.

but we can forecast other things without going into prophet mode. China has proven it can make things cheap, with reasonable quality, and at huge volume. and then, energy is getting cheap, solar prices are now lower than fossil fuel, the only limitation being dependent on sun. that means if there is demand for energy, the supply is now very elastic, as long as you are ok with pausing the activity during night and cloudy days (or doing it in a desert). this is totally different from the economics from the last century where energy was one of the main bottlenecks of economic development. yes the US is still fixated on oil, but then it’s run by dude born shortly after ww2 and their main issue is they’re scared of relying on chinese solar tech.

but globally speaking, energy supply elasticity means energy use will not be a bottleneck for economic development going forward. what else? human labor is getting automated, with or without AI. the automation trend has always been there. AI simply accelerates it. human labor will become cheap and abundant as any other economic goods. this is also a multi-decade trend.

i ranted about inflation (as CPI) being a weird metric because things are getting cheaper, making the CPI-inflation much lower than the money inflation. there are a couple things we’re talking about here: money is getting cheap fastest, goods are getting cheap less fast, and wages are getting cheap slightly slower than consumer goods.

money getting cheap may turn out to be an interesting thing. rapid inflation due to increased money supply might actually provide a soft landing for the systematic disruption due to abundance of everything. i’m now not very convinced that there will be a collapse of the fiat in the way like some currencies have blown up in the past – it’s more likely that there will be big inflation, but not hyper inflation where money becomes worth less than the paper it’s printed on, but rather, the value of the commodities will kind of follow along. value of money is relative to the things you can buy, so while systematically speaking the money may actually hyper-inflating, the abundance of everything (which includes money and goods) will make the relative-price stable enough so that it just feels like normal high inflation. in short, the money will become almost worthless, but there’s so much stuff being produced that it almost doesn’t matter.

i’ve also given this a bit of thought. like, in a world where commodities are dirt cheap, what happens if there is a worldwide collapse of the financial system? if the 20 largest banks in the world suddenly collapsed (let’s say they all signed a legally binding contract with me to buy my toenails for 100 trillion dollars), what would happen? well, the financial system would collapse, but like… would people actually suffer or die? i’m really not so sure. i think likely people would realize it’s all a game of monopoly (the board game) at this point, and maybe the main issues are land/property disputes that needed to be sorted out, because like, cheap stuff from china that costs almost nothing doesn’t mean i can live in your house. but like, food is almost free at this point and so are commodities, so why stress and worry?

another equivalent way to look at this is human labor cost (and demand). in theory economic cost all boils down to paying for human labor. the copper doesn’t need to get paid, but the copper miner needs to be paid. now we have copper mining robots, but the designers and workers who makes the robots need to get paid. but what if robots makes the robots? suppose in a benign world where they don’t become our overlords (i think these doomsday possibilities are overblown, i am constantly amazed at the number of people who are still mentally living in feudal times [aka 封建時代], eg. overlords are a very feudal concept), then no human actually needs to get paid. if this logic is sufficiently true (i.e. most work is done by robots and automated systems), then obviously money won’t be needed to buy anything because nobody is getting paid, and nobody has any money (except those who already have money. but money is useless now). money becomes obsolete.

there are probably exceptions. there are things that AI robots can’t do inherently. sure, some of the oft-repeated things like “creativity” is probably going to be mainly a human endeavor for a couple years, but i’m talking about fundamental things that no amount of intelligence functionally-equivalent to humans can mimick: real human demand.

Like, it all makes sense if you really think about it. In a world of post-scarcity, the only scarce thing is human demand. Without human needs and desires, the whole system is meaningless. Sure, maybe in 50 years the AI-robot-designed systems can pump out gigachad watts of free energy that powers data centers that can solve any conceivable problem within 3 seconds… but it’s all meaningless unless there’s a human that wants it. Remember we’re still talking about the benign world, no AI robot overlords. The purpose of the post-scarcity system is still to serve human needs. And thus what it cannot produce in over-abundance is human needs and desires, aka real human demand.

I’ve never subscribed to the idea that human “greed” is unlimited. In modern economic systems you can obviously make games where you want the number to go higher and higher… but that’s just a game. humans only play the numbers game when they perceive it to be meaningful, and in the old system the numbers are meaningful because the things are scarce. people have collections of luxury watches, handbags or clothes, they don’t have collections of wood chips, toilet paper, and cardboard boxes. and when everything is as cheap as toilet paper, people will realize their greed is so petty and easily defeated by their lack of imagination. greed is just another word for “demand”, and in a world of post-scarcity you will need a lot of imagination to actually want something that is not already cheaply available.

i know because i’m thinking hard about what people might want and come up blank. if i can predict what people will “greed” on, i can stack those things and become rich. but i’m coming up blank. so 40% it’s because i suck at this prophet game, and 60% it’s a real thing. in the transition period, sure, there will be things that will rise in price as a matter of fashionable trend, but generally things will just go down in price, and at the same time money will become cheaper even faster. but nobody will notice how fast it is because everything becomes cheap, and price is relative. the only thing they will feel the effect of is automation, making it harder to find paid work, and the lack of consumer urges.

it may be hard to imagine for some, but i can imagine a post-scarcity world where the whole economic game is flipped, and the system begs humans to give it something meaningful to do, because its whole purpose of existence is to meet human demands. paradoxically, being humans, the thing we will ask of it, is for a simulation game where there are no pesky AI robots asking me what I want [because I don’t know what else I want!] and instead just have a system tell me what to do so I don’t have to think about it. … …. … and BOOM! You’re back in 1999, baby. Welcome to the Matrix.

But are the doomsday scenarios more likely? Well, only if people lack imagination. There’s literally (almost) no reason to be hostile to another person if everything you can imagine you want is cheaply provided for. The only thing that a post-scarcity system cannot provide, besides real human demand, is extreme inequality. There can be some people who take offense at a system that does not withhold things from people they consider unworthy. People who derive pleasure only from having other people suffer. Those are the only enemies of the post-scarcity system. And only if they’re not satisfied with entering a 1999-era simulation and satisfying their kinks there and instead insist on doing it in the “real world” (* note that this world, the one in which you’re reading this message, is definitely simulated).

For this prophecy(?) (it feels more like a forecast at this point) I’ll probably be happy to realize I’m wrong because then I’ll know what to invest in over the long term. Like, sure, I can try to trade short/medium term, but if money becomes meaningless in 20 years I’d feel like an idiot. So prove me wrong.