> Why is this time different?
That depends if AI gets to the point where it can fully replace workers, as opposed to just augmenting them. I heard Alex Imas on a podcast recently talking about how a SWE can be running 10 agents to be 10x as productive, then that SWE is more valuable so firms should want to hire more SWEs and pay them more.
That works for a while, but what if AI gets to the point where it can manage the 10 agents as well as the SWE? Of course you could say the SWE can now manage 10 agents who each manage 10 agents so he's even more valuable, but that has to break down eventually. You don't need 1,000 SWEs each managing 10,000 agents - you hit a bottleneck in the ability to give them work fast enough (even if you need the SWE at the top at all).
I think it's easier to think of from the perspective of blue collar labor. It's further out there time-wise, but let's assume we get a humanoid robot that can do any labor a human can do. It costs $25,000 and maybe a couple grand a year to operate. Works 20 hours a day when it's not charging.
The construction worker it replaces isn't going to start managing a team of robots construction robots - there's already a GC doing that, and you can't scale building nearly the way you can scale writing code because of physical constraints. When the robot I've described exists, a huge swathe of the population is going to be unemployed. There's no competitor to hire them because the competitors just get robots too.
>The construction worker it replaces isn't going to start managing a team of robots construction robots
Some of them will. You've slashed construction labor cost to 5% of what it was before. With that and a similar reduction throughout the supply chain means we're going to start building a lot more stuff.
Even if we start building a lot more stuff, you still don't need those construction workers. You have a GC to manage the whole project, aided by an AI who's handling scheduling/operations/logistics. You have a detailed plan to build against.
So why do you need former construction workers to manage the robots? Why can't the GC and management AI run the whole thing?
Maybe there's some scenario where you still need something like one licensed person from each skilled trade to be responsible for the robots employing those trades. But there's no way you need everyone who worked on building sites managing robots, no matter how much construction you're doing.
That's pretty far away from where we're at. If things do get that far it's not going to be a problem. Eventually the robots will murder us in our sleep and our worries will be over.
For which poor unemployed people who just got laid off due to AI are the Robots building house for? More abstractly, for whom are we creating are these productivity miracles and surplus for. Does a rich person suddenly need a million iPhones for himself and himself alone?
[dead]
There's not going to be much additional poor unemployed people. They're going to get different jobs.
Doing what, exactly?
Could you have predicted in 1920 that in 2020 that 48% of the country would stop farming and some would be programmers?
"Past performance is not indicative of future results". People stopped being in farms because automation freed manual labor and people could move up in the value chain. We are getting to the point where there will be no chain left to go up to, at all. If you don't see the difference, I don't know what else to tell you.
this time we're automating humans that can do anything that humans can do. you should be asking what happened to the horses which were replaced by the tractors. the answer is elsewhere in the thread, their pop dropped 88% or w/e, we didnt need them anymore.
Labor costs are not the limiting factor on production.
If they were suddenly cut to near zero, a lot of projects that were previous uneconomical become viable.
"Becoming viable" does not mean "automatically put into execution". You still need to take overall demand in account.
Consider this: if demand was not a factor, anyone living in a moderately wealthy country would be practicing labor arbitrage and sending money to poorer places. Ask yourself why this doesn't happen.
I'm sorry, I must be missing your point, because isn't that exactly what's happening with manufacturing having gone to China?
No. It's different in two ways:
1) corporations moved manufacturing to where labor was cheap, but brought back the goods to sell them. This only works for as long as there is healhy consumer market somewhere. If AI really gets to automate most white-collar work, there will be no healthy consumer market left anywhere around the world.
2) The essay touches on this: any of the previous offshoring / job displacement movements happened on a much longer timeframe than what is being pushed now by the powers that be.
What is?
In theory breakeven demand, but ecosystens are basically economies, so termites are break even demand, and that's not good news.
Demand and primary resources go way higher on the list.
Demand and primary resources are effectively infinite.
>primary resources are effectively infinite
You just solved economics?
We need eugenics to weed these people out of the gene pool.
What?!
The primary resource needed to build a home is land. Do we have infinite availability of land in desirable areas to build on?
In a world where it’s dramatically cheaper to build infrastructure like roads, power, and plumbing, lots more land becomes desirable as a place to live.
Take Phoenix, for example, once air conditioning became cheap and pervasive.
> In a world where it’s dramatically cheaper to build infrastructure...
You can not make things dramatically cheaper by bringing only the cost of labor to zero. Your argument is circular!
The argument isn’t that AI brings the labor cost down to 0 in isolation. It brings the labor cost of the same amount of production down. So you get more production (more things = more supply -> lower prices) out of less labor.
> So you get more production out of less labor.
We don't need "AI" to figure out that technological advances increase productivity. The problem with yout argument is assuming that increased productivity mean overall reduced costs. It does not.
Healthcare, housing, education all have gone up despite increased productivity. Then you have things that are already so automated that there is no way to make them cheaper unless sacrificing quality - food, clothing, etc.
Then we have all the types of consumer products that have prices completely decoupled from the cost of labor. No one in their right mind the "cost of labor" has any relation whatsoever with Apple charging $1000 for an iPhone and/or Motorola charging $180 for a Moto G.
Healthcare, housing, education all have gone up despite increased productivity.
The hypothesis of Baumol’s Cost Disease is that these industries are exactly where we should expect prices to rise because they’re still dependent on low-productivity-growth human labor.
Baumol's Cost Disease is about relative costs. We are talking about productivity enabling absolute reduction in costs. Don't mix them up.
We were talking about infrastructure costs under increasing labor productivity. Now what are we talking about?
If the premise is that AI won’t improve productivity in industries like healthcare, education, and housing construction, then why are we worried about “the dead economy”?
No. You are getting it backwards. The premise is: even if AI improves productivity, we the people are not going to benefit from it.
The mistake you are making is that you are assuming that a system where productivity per unit of labor is higher automatically translates into increased global output. It does not. This idea of a dead economy theory is precisely the concern we are heading to a world where machines can make practically everything on the cheap, but it won't matter because the moneyed class won't need to satisfy the demands of the general populace.
So we have a bunch of billionaires sitting around, surveying a world where a much smaller amount of labor will produce a much larger amount of output, and they collectively decide not to hire that labor or spend capital on the technology that generates that output in combination with that labor because… they have enough money already?
No, ffs. You are missing that if they can do whatever they want with just "a small amount of labor", then the whole system gets to a point where Capital becomes the bottleneck for global productivity. People can not be trained faster than the machines can be created, so all that capital will go to an increasingly smaller number of workers.
To illustrate the point: Facebook laid off thousands of developers at the same time that it was hiring AI researchers, paying them tens of millions of dollars as a signing bonus.
Facebook (Meta) mostly “makes” ad space. So in that case, they’re making more / better ad space for the same inputs.
Online advertising is a competitive business, so that means more bang for the buck for Facebook’s advertisers. Now those advertisers have more money to invest in making more / better of whatever they make.
[dead]
Its also the 'labor theory of value'. That's the economic theory that Communism is based on. It has never been accurate and wasn't even considered legitimate during Marx's lifetime. It has possibly the worst track record of predictions of any theory ever conceived by people. Yet somehow academics still reference it. Nobody who actually is impacted by making the wrong economic predictions does though. Funny that...
Solve the diamond/water paradox?
Enough that it's effectively infinite, yes. Especially if we are imagining a world where subways cost 1/20th of what they do today.
> where subways cost 1/20th of what they do today.
1) We are talking about reducing the cost of labor, not overall costs.
2) Your logic only applies in the micro, not on the macro. If the cost of producing one thing goes down while population keeps their purchasing power, then what you are saying would make sense. The whole point of the article is that accelerated automation can bring a scenario where the cost of producing "things" would go down, but the economically active population would shrink drastically.
Someone will always have to prompt the AI, it can't just do that on its own. Or rather, maybe it can (you can just prompt it to "kindly do the needful" in a completely unspecific way) but the results won't be any good.
Sure, but the question is at what layer of abstraction do you have to prompt the AI?
You used to have to prompt the AI by starting to write the actual line of code you want, which it could autocomplete. Then you had to prompt it to write simple scripts or functions. The amount of scope you can prompt keeps getting bigger and bigger. Eventually, you have a PM or a CEO just telling it what features you need. Maybe it's a PM and a designer and a CEO and a CTO, but it will eventually get to the point where the number of people you need to do the prompting shrinks orders of magnitude from company sizes today. Maybe you just give the AI some money, prompt it to start a money-making business, and it goes out and does the same research and analysis that a seasoned entrepreneur would do to find an opportunity then builds out the business from there.
> the results won't be any good
Maybe, but I wouldn't bet on that. The trend over time has been that the results from prompting AI to do things have gotten better. I used to prompt it to build me dashboards and it would fail spectacularly. Now it one-shots them. Maybe the code is terrible (though doesn't matter for me, I'm the only one using it and I can verify the dashboard content is correct), but if the trend continues, it'll get better. Maybe the trend won't continue, but I've yet to come across a good explanation of why AI capabilities will just top out and cease improving forever.
Eventually all prompts distill to "raise the stock price year-over-year". Once we get CEO-bot, that's all we'll have to tell em.
Always? I mean, that's the hope, but they only have to be good enough to be of use for the rest of us to be unneeded.