DarkDude started this discussion 3 months ago#130,186
Japan used to be the second largest economy. Now they’re in 4th place and it’s basically inevitable they’re going to be surpassed by India in the near future and fall to the 5th place.
What went wrong? Other than not having kids and not having immigration. That’s probably like 90% of the reason, but I want to hear some other explanations from the people on here who are opposed to immigration who don’t have any kids.
These robots already exist, they just need a few years to start mass production.
Citizens having babies means waiting a decade and a half before they can begin basic employment.
What jobs do you think these people will be doing? Most jobs today can already be automated, theres just a delay while production ramps.
And no, 95.9% of people are not employed. About 38% of people are "not in the workforce". The number you gave it for people that are "in the workforce", people that are either already employed, or unemployed but still actively looking.
Add the officially unemployed to the "not in the workforce" and that's about 2 in 5 who arent working.
When 2 in 5 arent working, and we are a few years away from Ph.D level intelligences with advanced hands from stepping onto the shop floor... Why do we need more people? The native babies wont be ready before full automation arrives, and the immigrants are a short-term solution with long term consequences for culture and politics.
DarkDude joined in and replied with this 3 months ago, 16 minutes later, 16 hours after the original post[^][v]#1,397,494
Two more serious examples though would be Germany and the United States, Japan used to have a larger economy than Germany, but now Germany has a larger economy than Japan because German’s economy hasn’t declined like Japan’s economy, because their population hasn’t declined like Japan, because they accepted immigrants. The United States is also only 2.9% Native American, and the United States is the world’s largest economy despite having a population that’s 97.1% either immigrants, the descendants of immigrants, or the descendants of slaves, although calling slaves immigrants is controversial because that wasn’t really consensual.
> Will you answer the question? What jobs are those people born now going to do in 15 years?
There’s no point in arguing with someone who thinks AI is a magic technology because you’re just going to say that AI will improve exponentially (I have reason to believe that’s not physically possible), and become massively profitable. But right now it’s not profitable, and the funding is circular.
> > Will you answer the question? What jobs are those people born now going to do in 15 years? > > There’s no point in arguing
That's a cop-out for a very simple question.
> with someone who thinks AI is a magic technology because you’re just going to say that AI will improve exponentially (I have reason to believe that’s not physically possible), and become massively profitable. But right now it’s not profitable, and the funding is circular.
I didn't say any of that, and whether AI profitable to investors is a different matter than whether technology can automate a job.
DarkDude joined in and replied with this 3 months ago, 9 minutes later, 17 hours after the original post[^][v]#1,397,507
@previous (B)
Someone has to pay for all those GPUs. Electricity isn’t free. They aren’t generating more revenue than it takes to maintain the infrastructure. Either it’s viable or it’s not viable. It’s not that complicated.
You're still confusing two different ideas, even though I just explained the difference:
A) Whether an AI stock is a good investment, whether it can meet expected ROI
B) Whether a technology will be adopted to replace labor, regardless of whether the company providing it has small or big profit margins.
There was a .com bubble that popped in 2000. What happened to the web in the years after? Broadband penetration grew rapidly, and people used the web significantly more. You seem completely unable to tell the difference between these two concepts. Even though expected profits werent there, people still used the technology more and more. Part of the reason for both was the that competition in that industry pushed down consumer prices, which means investors lost and consumers had more reason to use it.
If you have to keep avoiding the question, it's obvious you don't know what you're talking about. You said we need workers, but you can't name a single thing those workers are needed for? Really?
DarkDude replied with this 3 months ago, 5 minutes later, 17 hours after the original post[^][v]#1,397,509
@previous (B)
Nvidia invests in OpenAI so that OpenAI can buy chips from Nvidia, OpenAI isn’t profitable.
If I want to use ChatGPT to replace a programmer, and I’m paying OpenAI less than it costs them to pay for their GPUs and electricity, and OpenAI gets paid by Nvidia to buy Nvidia GPUs, and Nvidia gets their revenue by selling GPUs to OpenAI, do you not see how it doesn’t matter whether or not ChatGPT can replace a programmer?
Anonymous B replied with this 3 months ago, 31 minutes later, 18 hours after the original post[^][v]#1,397,515
@previous (DarkDude)
You can run models that write code on a gaming PC today. Renting a server for the flagship models, without any subsidies is very cheap.
If the tech stopped advancing today, there would still be a wave of automation.
DarkDude replied with this 3 months ago, 6 minutes later, 18 hours after the original post[^][v]#1,397,521
@previous (B)
That’s dependent upon knowing things that are unknowable. The most common jobs are rather boring ordinary things like construction worker, teacher, cashier, electrician, mechanic, bartender, etc. The correct answer would be something of them but possibly not all of them. Which ones specifically will still be around and which ones specifically won’t, if I start guessing at that, the probability that I’m wrong will get a lot higher. I can make general statements, but being overly specific is too risky.
DarkDude double-posted this 3 months ago, 1 minute later, 18 hours after the original post[^][v]#1,397,522
So if I had to make a guess, I would say that 50 years from now, out of these jobs: electrician, teacher, construction worker, cashier, bartender, mechanic, there people will still be doing at least 3 of those jobs. Which ones, I don’t know, all of them, possibly, none of them, highly unlikely.
DarkDude triple-posted this 3 months ago, 5 minutes later, 18 hours after the original post[^][v]#1,397,523
It’s also not really clear what it means whether "people do a job." For example, self checkout has been a thing all my life, but for some reason half the time they pay some guy to stand there to scan my groceries for some reason. So does a job not existing mean half as many people do it, does it mean zero people do it, it’s a little bit vaguely defined.
If we have AI that can talk like a professional in those fields, can recognize whats happening in a visual feed, and we have robots with hands...
Why would a human do any of the things you mentioned? The electricity for a robot is nothing compared to salary + taxes + insurance + liability for a human.
Those jobs would only exist if the government mandated a human do it, which is the opposite situation of a society that needs more births and immigrants to fill jobs.
DarkDude replied with this 3 months ago, 4 minutes later, 19 hours after the original post[^][v]#1,397,526
@1,397,524 (B)
I don’t see how "talking like a professional in the field" is impressive. If you train a large language model on scientific papers and then you ask it to spit out something that looks like a scientific paper and it spits out something that looks like a scientific paper because it knows what words in what order are correlated with sounding like a scientific paper, I’d hardly call that intelligence.
The cost of electricity isn’t negligible. Humans are actually very energy efficient.
> I don’t see how "talking like a professional in the field" is impressive. If you train a large language model on scientific papers and then you ask it to spit out something that looks like a scientific paper and it spits out something that looks like a scientific paper because it knows what words in what order are correlated with sounding like a scientific paper, I’d hardly call that intelligence.
I didn't say you should be impressed, I said it could automate jobs.
If you can make a solid philosophical argument for why its not really thinking, does that change the fact the bot replaced a human in a job?
> The cost of electricity isn’t negligible. Humans are actually very energy efficient.
No, lol, not even close. It doesnt cost $50,000 to run a GPU all year, and the humanoid bots that exist don't cost tens of thousands in emergy expenses.
DarkDude replied with this 3 months ago, 3 minutes later, 22 hours after the original post[^][v]#1,397,548
@previous (B)
I’m going to be honest: I feel highly skeptical of the utility of humanoid robots. In a manufacturing setting, I would think that robots specialized to the task at hand would be more efficient than humanoid robots, and in jobs that require mental labor, there isn’t really a point to having a physical robot in the first place.
DarkDude double-posted this 3 months ago, 7 minutes later, 22 hours after the original post[^][v]#1,397,549
For example, how would taking a machine like this one that is used in manufacturing smartphone batteries, and replacing it with more expensive humanoid robots that are less specialized and more difficult to program be a better solution?
Humanoid robots arent a replacement for specialized machines, they are a replacement for jobs that can't be done by specialized machines and are done by people.