Gendai Shakai de Otome Game no Akuyaku Reijou wo Suru no wa Chotto Taihen - Ch. 22 - The Lost Generation

[MangaDex]

 

[sathornet]

We got the AI ice age now

History do rhythms

 

[ciera]

Thank you for the update

 

[Haezer]

This manga is supposed to be a work of fiction

 

[SiliconSignatureSwissChair]

We got the AI ice age now

History do rhythms

AI, in it's current form, is simply a convenient alibi for cost cutting layoffs.

 

[polartrain]

Man you guys are my favourite scanlation group. Your notes are also great every time, though this time it hits a little close to home (cs despair).

 

[Shiden]

On the Runa-sensei page, I think that was supposed to be mid-1990s, rather than the mid-1890s, otherwise that's one heck of a frozen generation.

And... yeah. +1 to the comments on the last page.

 

[TheDarkStar]

Real rough chapter. Much easier to deal with a horde of demons than realistic despair. Sad that Runa doesn’t feel like she’s accomplished anything this second time around. Good thing there’s hope on the horizon.

 

[eng1]

AI, in it's current form, is simply a convenient alibi for cost cutting layoffs.

You underestimate the impact of AI in current form, yes the AI is dumb now and due to limitations is unlikely to have much improvements for a while.

But the true value of AI isn't it directly replacing people, but in same way automation does, as an assistant it can make it that 1 person can do the job of 3.

So jobs are going to be lost, at least until new industries become a thing to fill in the gaps.

 

[Slingerman]

I have a feeling the writer entered the Japanese workforce in the 1990s. Just a feeling!

Thanks for the scanlation!

 

[cryum]

Come to think of it, I haven't seen recent data. Is all this AI pandering going to result in another plunge in birth rates? When we're not even done with the first one?

 

[MoogleJaws]

Yeah, the lost generation hits a little close to home. There are more jobs than ever because we're all working three of them just to get by.

 

[DoingScience]

You underestimate the impact of AI in current form, yes the AI is dumb now and due to limitations is unlikely to have much improvements for a while.

It's also important to remember that AI goes beyond chat bots. There were neural networks that can identify cancer before a human could hope to years ago.

But also

 

[eng1]

It's also important to remember that AI goes beyond chat bots. There were neural networks that can identify cancer before a human could hope to years ago.

But also

The real problem is the training data. There was an attempt to have ai identity cancer, after training the AI, it concluded that if there is a ruler in the picture, it is cancer.

So simply put they can assist but current AI simply can't be relied on. And we have reached a point where pretty much all public data is already been used on AI training. And worse, most new data on the Internet is AI made. So the end result is we are getting in a loop where AI data is being poisoned by new low quality data by other AIs.

And while sure AI maybe able to help find cancer, there is also the opposite situation where people rely on AI for diagnostics instead of going to doctors or even doctors themselves getting lazy and going off what AI gives them.

Then there is a huge risk of us losing the Internet as we know it, watch as AI bots out number humans. The whole Internet will turn into ai bots posting content, ai bots responding to content and other AI bots responding back to them. It's gonna be a total nightmare where everything will go behind paywalls

 

[Liar]

It's so hard to find a job these days.

 

[mrpointy]

[...]
But the true value of AI isn't it directly replacing people, but in same way automation does, as an assistant it can make it that 1 person can do the job of 3.
[...]

If that were actually the case, the stories would all be "we can do 3 times more for the same labor cost, productivity through the roof, rah rah rah". They're not, so someone's lying.

Instead, it's as the OP said: suppose you're a CEO and business is not going well and you want to fire folks to cut cost. You could do that as is, but then the market knows business is not going well, the stock drops, and so your compensation drops. Bad for you. So instead, you make up a story about productivity gains through AI making people redundant. The end result is the same: you still fire folks to cut costs, but now your peers call you a visionary, the stock goes up, you can buy another yacht and life is good.

 

[Wolvenworks]

TL;DR it sucks to be a millennial if you happen to have a particular set of skills that are write-off-able. I’ve heard of a case where someone with a master degree in math ended up being an office boy…companies may say “we don’t discriminate in hiring”, but internationally-speaking, unless you’re SUPER GOOD that they HAVE to ship you out, they won’t consider you as much as local talent, which sucks when you’re in Indonesia and the paycheck’s more likely to go anywhere but up.

It double sucks when your degree doesn’t quite line up with your work experience and certifications. Domestically, they don’t quite need you and your niche skillset. Internationally, you’re a nobody…

 

[sdarkpaladin]

I feel that the problem right now is that AI is a solution looking for a problem, which would be good if not for the fact that they literally cost humoungous amount of resources.

Unlike the dot com bubble or the Sub-prime mortgage crisis, there's two different counters ticking down.

Not only is there one for total investment, there's also one for total resources required.

Everybody can host their own server at home and make a webpage.

A small team can definitely build a house.

But you need investments in the millions to even run an AI locally.

So I think that bubble will burst faster than the other bubbles.

The problem is where does that leave you (or me)? Will we be capitalizing on the bubble bursting? Or will we be part of the collateral damage.

 

[eng1]

If that were actually the case, the stories would all be "we can do 3 times more for the same labor cost, productivity through the roof, rah rah rah". They're not, so someone's lying.

Instead, it's as the OP said: suppose you're a CEO and business is not going well and you want to fire folks to cut cost. You could do that as is, but then the market knows business is not going well, the stock drops, and so your compensation drops. Bad for you. So instead, you make up a story about productivity gains through AI making people redundant. The end result is the same: you still fire folks to cut costs, but now your peers call you a visionary, the stock goes up, you can buy another yacht and life is good.

We are still in early days of AI adoption in workflow, so you aren't going to see huge jumps in productivity. Especially since many large businesses overhired to facilitate growth but don't need them anymore. Even businesses who are doing well financially are firing people due to AI.

If they wanted an excuse to fire, the can do what they always did. Hire a consulting agency who will always recommend firing and blame the agency.

I feel that the problem right now is that AI is a solution looking for a problem, which would be good if not for the fact that they literally cost humoungous amount of resources.

Unlike the dot com bubble or the Sub-prime mortgage crisis, there's two different counters ticking down.

Not only is there one for total investment, there's also one for total resources required.

Everybody can host their own server at home and make a webpage.

A small team can definitely build a house.

But you need investments in the millions to even run an AI locally.

So I think that bubble will burst faster than the other bubbles.

The problem is where does that leave you (or me)? Will we be capitalizing on the bubble bursting? Or will we be part of the collateral damage.

You don't need an investment in the millions to run AI locally. Your PC is already likely doing AI locally, just usually simple tasks like grammar checking, code completion, or translating.

A gaming pc can also run many LLMs, a regular 12gb vram GPU can run 8B-12B quant model.

You can even run some of the big models locally by getting a 128gb and stricks halo.

The only big problem for locally running AI is the need for fast ram. And due to the AI craze, the large companies reserved all the ram. So ram prices are going through the roof.