>I don't think it's a matter of foresight.

What you describe thereafter is exactly foresight, just not on an individual scale. Governmental foresight with implementing security nets.

The US has mega rich corporations, but a lot of countries don't. However the US also has a pretty large population compared to other fully developed countries. Social security has been the target for stripping down over the years, and with the generation currently reaping its benefits, projections show younger generations will be with less. But that's more political than I care to dive into. And may not be the case in the US, I'm not from the states I'm north of the border.

I think cracks will form though, sure we have systems for unemployment, but those systems haven't been tested for crises levels of unemployment. It also begs the question of UBI being available, while some people continue to work.

I have no doubt we will make it, it's just a matter of how much foresight we have to reduce suffering among the people that don't make the decisions.

Meanwhile jobs are being automated already.

Part of me is anxious about automation affecting me, but a larger part of me is anxious in how the general population will react if their oblivious world is shattered. In theory it sounds great saying "I told you so" but that's never as sweet as it's imagined when people are frustrated and suffering.

I can't think of a better solution. I don't know how UBI would be implemented, especially in countries without giant trillion dollar corporations.

I think before even AGI we will be caught off guard. I think 2023 language models will surprise us.

I don't even talk about the singularity really. People can't even talk about the next few years of language models. Singularity is way too sci-fi sounding to them.

People have done it, I recall some posts about it. I'm not able to verify the accuracy of those claims, but why wouldn't it work? If you include the ability to edit it in your experiment then it's just doing the typing for you. You still input prompts with your ideas.

The next generation will certainly be better for cutting down on the need to edit the essay yourself too. Hopefully those are released before you finish school.

That's my fear too. It's what I try to talk about, but get met with either people agreeing and having no idea if a solution (same boat), or people on the total end of the spectrum saying it isn't happening there will always be enough jobs.

I don't care about people's prediction of when they think it will happen, More about when it happens what will we do in that period of transition.

I made the askreddit thread to see what jobs people think won't be replaced. Some of the answers are good candidates for lasting a while on the physical side of their job, but most of them could be fully automated outside of the physical aspects. What's interesting is one person even supposing a robot existed, thinks firefighting will never be automated because you need a gut feeling and have hair stand up on the back of your neck. Another saying mixing music needs a human ear. It's a case of not understanding humans won't be the most intelligent thing on the planet for long, and our senses can be replicated and improved magnitudes over. Intuition isn't uniquely human, and it isn't magic.

I guess I have been asking the wrong question and should phrase it as a hypothetical more, like "if half of all jobs were automated, what would society do" or something.

>Just imagine 25 years ago, could you have predicted the explosion of work related to the internet?

We knew it would be big, we weren't sure how. But with automation, specifically targeting jobs that are mostly done over the internet first, it isn't especially difficult to imagine those jobs in tech being replaced by new tech. Evolution of tech companies.

If we look at it from the perspective of corporations instead of individuals with morals, it makes sense for companies to want to do these "new" high paying jobs with automation. The age of AI will dwarf the age of the internet. It's not really a good comparison.

>How is that logic "automating even part of a job leads to layoffs" standing up to the test?

We haven't had automation that can do every aspect of a job better than a human. I'm not trying to convince people to join me in panic, I have some anxiety about it, but the absolute disregard and "everything will be fine and life will go on the way it always has" is not a productive mindset. I'm asking you, call it hypothetically if you want, what if you're wrong? Are you so confident you haven't given any other option even a moment of thought?

>Our desires scale up faster than automation or resources.

Except it won't if automation starts scaling anywhere near the rate of technological innovation.

Look, I realize there's those of us that probably worry too much and we sound crazy, but the majority of people probably don't worry enough. You can call it being grounded, but I can call it being unprepared. Even if it's somewhere in the middle, which it hopefully will be, do you personally have any ideas for the mere possibility? What would the best course of action be if a significant amount of people are unemployed? If your job is safe, do you think it's fair if a UBI exists? Would you take UBI if your job was replaced and you had to feed your family? Can you even imagine a world post-scarcity and people not working?

There's a pretty quick diminishing return on lower prices increasing demand. Razor thin margins with high volume usually leads to price increases until you maintain the perfect balance for profit.

There's just not an unlimited output that companies can produce and still be successful, especially with thin margins.

Capitalism chases quarterly profits, you lay people off to balance output to sales.

>Something like everyone having to make the neighborhood more beautiful and clean and nice to live in 8 hours a day.

In your opinion, how many hours a day/week do you imagine people doing this? I know Japanese work culture is much different than the west. What in your view is the "right" amount of hours to work to qualify as not being perceived as lazy?

>Which makes me think your one of those folks OP called out for just not understanding the concept of Singularity or AI development in general.

I never thought about that happening, my bad.

I honestly have no clue. Predicting the economy is impossible, especially when you can't really predict the variables. I'm looking for an answer that makes sense instead of the same ideas that don't.

It doesn't make sense to increase productivity and output in tandem with humans infinitely. Not only will AI have much higher intellect than humans, most humans won't be able to do highly intellectual jobs. So that leaves labour and skilled positions that AI haven't been able to automate and only a handful of people can do. The labour careers won't have high salaries, especially if they could be replaced with automated systems but it's just due to cost.

So we will live in a world where the majority is working low wage, unskilled jobs while companies produce an over-abundance of products that people can't afford? Who will their customers be?

You could argue automation would decrease prices, but that decreases prices of robots because they're essentially a product to be produced, and that makes low skilled labourers' cost more than automation.

I haven't heard much that's more convincing than just enjoying leisure time, even though that feels like a pipe dream.

That was what surprised me. And the amount of people that say "never" as in not even 100 years, just NEVER.

Not fully automated, but when the time comes and 2 AI can replace 15 people because so much manpower is spent on tasks that can be automated easily I think we will see corporations adopting that in droves.

Does that leave enough work for the majority though? Or does it leave a case of haves and have-nots?

Only for so long though. You can't just keep increasing output.

So you think companies will choose to increase output infinitely while AI improves, rather than save money on wages? There's only so much productivity to be had when something you're selling has a limited customer base. It'd be a weird world where we all still have to work even though we have an over abundance of material goods, living standards, and wealth.

I'd argue at least a large portion of people will very likely lose their current jobs to automation. While no, I can't promise that, it's hard to imagine a world where we just stop using language models or advancing them / something even better.

I don't think wages will remain the same, and ever be cheaper than a corporation buying robots in bulk and installing their own AI software in them. What about when material gathering and production is automated? All technology has decreased in price over time when you take into account the efficiency and power of it.

I think roll out is more a matter of time than an "if" because at some point corporations will chase the profit, and others will fall far behind.

>willing to create a system of bullshit jobs just to force people to "work".

Yeah I'm really hoping that doesn't become the case. That's what it seems like everyone thinks is ideal too. Like automate everything, but keep people doing meaningless work just to feel like they have a purpose.

That and they can't seem to comprehend humans aren't special when it comes to things like creativity, fixing things, customer service/relations, etc. There's a lot of people that say "never" and that AI has been touted as automating jobs for decades. I guess they never thought eventually that decade would come.

I did that for years too. But I just can't stop thinking about it these days, probably because of the constant articles about new scaling and breakthroughs.

I think everyone saw the text to image stuff, and kind of just figured it'd be years before that ever got advanced enough to replace anyone. The graphic design subreddit had some posts about it and most replies there were along the lines of "it's a great productivity tool" with only the occasional "time to find a new career."

Was the job very text based? I'm surprised if not, most people associate AI today as text only, or drawing pictures.