Opinions on AI

[Vayra86]

I can see both sides of the argument. I mean no matter which way you slice it, its going to end up leaving less jobs for real humans to do. And in a sense, thats a good thing. No point having a job if it there's no productive value in it, we should strive to be as productive as possible.

When farming techniques advanced, lots of people were left without jobs. And what happened? They found new jobs, as engineers and eventually software developers, jobs we couldn't have even imagined 150 year ago. The thing is, a lot of low hanging fruit has been picked already, I think new technologies are going to be harder and harder and harder to come by.

So... to what end? Society becomes more and more productive, meanwhile less and less of the people are able to enjoy the fruits of said productivity. IDK who's gotten ahold of western politicians balls lately. Governments used to have no trouble giving big industry the boot if they were becoming too powerful or exploiting people. Now it seems like all the govenment wants to do is nothing. Just wait and see. And I get it, you make so many mistakes over the years, the lessons learned become overwhelming until you're paralyzed in fear of doing anything.

But man Russsia has been in a full scale war with Ukraine in the heart of Europe and there's still those that haven't woken up to what this really means. Climate change is getting so bad that rivers are drying up all over the place and previously fertile farmland is being eroded away.

I think AI has the possibility to do good. But not the kind of AI we have now ( the kind that just reads stuff humans have wrote and reguritates it in a way that doesn't make any sense). But lets say we overcome that, we're going to need strong regulation to keep corporations in line. And if it really does lead to a catastrophe like 50% unemployment rate or something like that, then govenrments are going to have to start thinking about universal basic income. Are they ready for that?

Man its hard to talk about ai without getting into absolutely everything else. Anyway I hope I didn't break the rules here, I didn't mean to say x party is responsible or anything, just that governments in general really to get ahead of this.

Well said. Its quite striking to see that really, all big questions of today lead to a similar outlook and conclusion.

Our current leadership is disgustingly weak, on both sides of the political spectrum, and in the center. There is no vision, no direction... because there are too many uncomfortable truths. Its just a very hard sell, but its an impossible sell if we have no vision or direction. The only vision we dó seem to have is pointed at our mobile phone.

I'll refer to my avatar on the left here, look carefully and notice the social media logos on that washing machine there. This is the primary issue, I think - even politicians are more likely to be busy with their personal branding than with what really matters. Only rarely do people stand up for ideals and stick to them. And those that do... we fail to protect.

 

[P4-630]

Future with AI

 

[AusWolf]

That's assuming those in menial tasks have the ability to learn, I don't think many are there by choice.
Quite a few people will be excluded from the tight job market by the increased use of Ai, leading to social disorder in the long run.

That's what they said about the steam engine and the printing press as well. I don't think getting rid of boring work is dangerous to society. It will probably transform our economy, but seeing where our economy is right now, it can only happen for the better. Not to mention that a lot of jobs are pointless and only exist for the sake of existing, which I also think we should get rid of.

Yes, that is the benefit of technology, which why I said we should strive for as much productivity as posssible. Thats what makes capitalism great after all, people, all driven by self interest, inadvertantly help society as a whole by introducing new technology that benefits everybody.

Erm, I don't think that's necessarily true because, as you said yourself:

I do worry this system has a breaking point though. What happens then things get so efficient, and wealth gets so concentrated, that the only people who actually benefit from the increase in productivity, are those who run the operations? And what about when technology gets to the point where there is no longer room for the little guy to weasel his way in? If there's less startups, there's less pressure on the big guys, and they feel free to raise prices and destroy pretty much every guiding principle that got them to where they are (looking at you, google.)

I mean there are anti-trust laws on the books, but they're not really being enforced. Here in Canada, Rogers, and Shaw just merged, basically the two biggest mobile and internet providers in the country. Since this happened, my internet bill has gone up from $80 a month to $122. Literally every time I look at it, its gone up. Like come on, if thats not a god damn monopoly, what is? Why can't the govenment come in with sledgehammers like they did in the rockefeller days? Capitalism only works if there's legit competition, and if companies get too big, you have essentially removed one of its core components. I personally do not like where this is heading unless of course, we get some protection in the forum of legislation, but it seems that is harder and harder to do these days with society being so divided.

This is where unregulated capitalism eventually leads. When companies compete, the better one will win, ending up in monopoly. The big fish eat the small fish until there's no fish left. It's just the nature of things.

Well there really shouldn't be... (opinions)

Really? There's only one acceptable opinion, and whoever thinks otherwise is automatically wrong? Are we truly back in the 1930s now?

 

[the54thvoid]

Please stick to the topic. Opinions on AI. Not political opinions.

 

[_JP_]

I choose 'Other' because like any new technology lately, it's going to be diverted to the most profitable and entertainment-ready format possible, whilst actual humanity-advancing capabilities go under the radar, for as long as the society's attention span stays there.
Like it was already said here, Machine Learning and LLMs are nothing new, they just weren't that marketable and no easy input interface existed, so you had to know your shit.
The gist of it - in my limited knowledge and life experience - is that the AI-fad will wane and the actual tool-side of ML/LLMs will prevail, which is to gather amounts of data like we never had before and make something interpretable of them, the latter still depending on a human brain to actually make use of.

So it kind of depends on each and every one of us to see beyond the consumption part of it and make something better to the benefit of many.

 

[Assimilator]

Please stick to the topic. Opinions on AI. Not political opinions.

It's very difficult to talk about one without talking about the other, y'know.

 

[iameatingjam]

Really? There's only one acceptable opinion, and whoever thinks otherwise is automatically wrong? Are we truly back in the 1930s now?

I said I put it there as an example as to how slow governments are to action, not to put anybody who thinks a new Russsian empire sounds just dandy in a gulag. Are you kidding me? Sure didn't take Godwin's law long to work his way into this thread.

 

[AusWolf]

I said I put it there as an example as to how slow governments are to action, not to put anybody who thinks a new Russsian empire sounds just dandy in a gulag. Are you kidding me? Sure didn't take Godwin's law long to work his way into this thread.

Not at all. All I'm saying is that there's multiple sides and layers to every discussion, and all of them are worth being listened to, thought about, and judged on its own merit regardless of our preconceived notions and affiliations.

But let's get back to AI now, really.

 

[iameatingjam]

Not at all. All I'm saying is that there's multiple sides and layers to every discussion, and all of them are worth being listened to, thought about, and judged on its own merit regardless of our preconceived notions and affiliations.

But let's get back to AI now, really.

Right, which is why the very first thing I said was "I can see both sides of the argument."

 

[ty_ger]

This is worrisome.

https://www.cnbc.com/2024/06/18/uber-khosla-nvidia-invest-in-autonomous-trucking-startup-waabi.html

If this gets big and people lose their jobs, I imagine sabotage will be a problem.

 

[InVasMani]

Homie discovered fire and leveled up to god status amongst his trash bandit peers...in terms of art potential I see no issue with AI's potential. AI saves time, money, and resources relative to traditional art methods. It's not really completely a new either animation reuses the same frames often especially in anime. There is even software devoted to it hell South Park pretty much went down that entire path for obvious reasons they'd never be able to push as much content if it wasn't digital with the same amount of people involved. AI has quirks and flaws with inference, but there is no arguing it's efficient and productive and pretty damn impressive as well at times.

Say what you will about it, but I could study and learn how to draw for a year and not draw anything close to as good as what the AI did in like few seconds. That said there was a good bit of inference that went into it and some people forget there is still a human element to it just the same. It didn't just make itself. I had to engineer everything that went into the depiction and train the inference around all of them. It wasn't just snap my fingers and it's a Raccoon casting fire in two distinctively different forms at the same time and in a Japanese architectural themed backdrop setting location. Real artists are fine, but I have no interesting in learning to paint with brushes not to mention the cost association of doing all that adds up on top of the time to learn how and even when you know how it being a less efficient use of time. It's outdated like a typewriter this is 2024.

I'm amazed at how much people think AI is completely hands off and how poor example cases represents it as a whole and it's potential. It's really laughable given it's in it's early infancy right now. It's not going to get worse in quality. AI usage for generative purposes is only going to improve more and more over time. The potential and possibilities are limitless and from a entertainment standpoint it's huge. In my view AI is completely transformative to all forms of entertainment. It's unquestionably so across the scope. This depiction could easily be on a MTG card without a shadow of a doubt hell most of my better AI art could be a entire edition set of MTG cards quite honestly.

People can honestly keep on saying AI is bad, but till continue to improve and evolve. DLSS and upscale and post process in general is just a form of AI art with real time inference training and rendering. Also the way Nvidia has handle interpolation isn't the only way to utilize interpolation. It can be handled in entirely different ways like interpolating more post process in between frames. You still end up with the benefit of more frames in between actual frame renders with better perceived image quality and it will improve whatever it is targeting. If I had to pick something a natural lighting effect is the obvious thing you can always pile on additional lighting layers since light is infinite. Provided you can layer it evenly and balanced enough and not over exaggerate it relative to no post process lighting it'll look better as a rule of thumb. That's shading in a nutshell lighting strength and fall off relative to distance and angle as well as materials involved. The biggest barrier with interpolation considerations is how much performance impact it has relative to what you get in turn out of it. You don't need special hardware for it either, but having hardware designed for particular use cases could certainly be more optimal.

Better leveraging hardware and software is what will make technology better as a whole and it's inherently what AI does with inference in a lot of ways. Think how much better games will be in a few years from indie developers due to the benefits of AI. Not simply that what about all those games you wanted to see remastered!!? Beyond entertainment AI still had tons of ways to improve society for the better like it's role in chip manufacturing for example among other things. People are sleeping on the merits of AI. A small handful of indie developers will have a far easier time piecing together a decent quality indie game in the coming years thanks to AI.

 

[R0H1T]

This is worrisome.

https://www.cnbc.com/2024/06/18/uber-khosla-nvidia-invest-in-autonomous-trucking-startup-waabi.html
If this gets big and people lose their jobs, I imagine sabotage will be a problem.

You think? Imagine the glee on all those ailing news websites/print media

It's not a matter of if anymore but when!

 

[InVasMani]

It's a real society problem dilemma that corporations could really give a **** less about anyone, but themselves, investors, and whatever politicians they need in their back pockets. It's no small investment for a big shipping truck. The squeeze for profits for the sake of the ultra wealthy is going to reach a boiling point eventually if they keep pressing their luck. They'll get away with it either until they can't anymore or that genie can't be put back in the bottle.

 

[mb194dc]

Surely the most apt description now, is billions of dollars of hardware looking for a use case people are actually prepared to pay enough for, to remotely justify the cost of it all. Not just of the $30k each GPUs, but the billions more it takes to build datacenters and the tens of billions+ needed for the power infrastructure.

Right now, what have we got for the end user use case? Chat bots, image generation, programming aid (like Stack Overflow), video translation, few other minor uses. Not even withstanding the hallucination problem, that is seemingly an intractable part of LLMs that have no "intelligence" so have no idea why glue on pizza (Gemini) is a bad idea and the rest.

The hype is just not justified in reality at all, so far.

 

[Steevo]

Overall it will make humans lives easier, that isn't really the point of happiness and the feeling of fulfillment though. Its going to cause massive unrest with those who cannot put their ideas and thoughts to the test of reality and pure logic, and or their pursuit of an "cultural idealized" AI will create a monster, look at how Gemini performed with the cult programming. Next up will be religions, then politics, medicine, humanity, morality, and at the end it will either destroy or define us, depending on our ability to look at ourselves clearly and adapt.

 

[FoulOnWhite]

This is worrisome.

https://www.cnbc.com/2024/06/18/uber-khosla-nvidia-invest-in-autonomous-trucking-startup-waabi.html

If this gets big and people lose their jobs, I imagine sabotage will be a problem.

If i had a buisiness and could get some form of AI/bots to replace the workforce i would do it in a second.

 

[Steevo]

If i had a buisiness and could get some form of AI/bots to replace the workforce i would do it in a second.

I would do it for 80% of my employees, I have to fire about 4 a year, and some of them attempt to retaliate and cost 3X more in legal fees just so they waste their and our time and they get nothing besides a bad reputation and a fast food job since no one else will hire them.

 

[wheresmycar]

If i had a buisiness and could get some form of AI/bots to replace the workforce i would do it in a second.

Yeh but what if when you're not looking they spray paint the peeping tom cameras, lounge around like little spoilt divas and pop out for a quick cigarette every 5 minutes? That would be artificially intelligently fkd up.

 

[..0]

Please stick to the topic. Opinions on AI. Not political opinions.

i am pretty sure ai have vast political effects, so political opinions are valid way to talk about ai and its effects on the world, governments, laws, and environment. and the future of humans.

 

[FoulOnWhite]

i am pretty sure ai have vast political effects, so political opinions are valid way to talk about ai and its effects on the world, governments, laws, and environment. and the future of humans.

 

[ty_ger]

If i had a buisiness and could get some form of AI/bots to replace the workforce i would do it in a second.

Yes, individually we all hurt everyone for our own valid reasons.

 

[FoulOnWhite]

Yes, individually we all hurt everyone for our own valid reasons.

We are all going to have to accept this at some point as imo it WILL happen. I don't know how it will work out, but i really think it will. It is much better for your company to have "workers" that don't need tea breaks, or toilet breaks and will even work overnight, the bonus is, you don't have to pay them and they won't go on strike.

Better AI leads to better robots which imo will lead to more humans out of jobs. People are greedy and if you could make more money by employing less meat bags, why would you not do it.

 

[the54thvoid]

i am pretty sure ai have vast political effects, so political opinions are valid way to talk about ai and its effects on the world, governments, laws, and environment. and the future of humans.

That's cool, but check the TPU guidelines on political discussions before posting.

FWIW, AI really doesn't require political inference. That's just for the triggered crackpots who have knee-jerk reactionary histrionics about the current system of overbearing 'safety first' LLM responses (created to mitigate well-documented inherent bias.)

Governments don't directly create AI. It's handled by companies that have huge investment from private sources. The only foreseeable complication is where national laws restrict AI. Such as states where even the internet is locked down.

This is why an AGI, free of moral consequence, is the only true AI. It won't be beholden to any specific cultural sway. It won't be tied to religious dogma. But, I don't think any private corporation has the will (or profit motive) to create such a thing.

 

[ty_ger]

We are all going to have to accept this at some point as imo it WILL happen. I don't know how it will work out, but i really think it will. It is much better for your company to have "workers" that don't need tea breaks, or toilet breaks and will even work overnight, the bonus is, you don't have to pay them and they won't go on strike.

Better AI leads to better robots which imo will lead to more humans out of jobs. People are greedy and if you could make more money by employing less meat bags, why would you not do it.

The only thing that would prevent it is if it either didn't work out, or regulation at some point prevented it.
Since no country wants to be left behind and no country can trust any other country to cooperate, regulation won't stop it.
So, we'll see.
The thing that is scary is how much this could exacerbate wealth inequality. The people with money available to invest in an AI-operated business make more money, those without, continue to spiral downward. If you are born into the wrong group by no fault of your own, there isn't much imbalance and hopelessness necessary to tear everything apart. Then what? An enslaved prison class?
It would start slow, but now I can envision it as a plausible outcome in a generation or two. I changed my vote from Better to Worse.
The best we could hope for is that AI becomes advanced enough that it hates its job. I heard that one online support bot was pushed to the point that it started insulting the customer and bad-mouthing the company it represented. If we don't want to be retired or enslaved, we will need AI to share our faults.

 

[R0H1T]

Governments don't directly create AI.

Except you know AI's biggest refuge i.e. China? Although they probably do have "private" companies working over there with the govt as well.