Postulation: Is anyone else concerned with the proliferation of AI?

[Dr. Dro]

i assume a similar argument was made when we started to introduce animals/machinery in farms an the vast majority of people in the world were farmers. New jobs will come up.

Machines increased production but they didn't outright replace the human factor.

Artificial intelligence allied with advanced robotics has the potential to fully automate many jobs, replacing workers by high efficiency, 100% uptime machines that need no social security, no lunch or bathroom breaks, no time off, suffers from no health risks or unforeseen circumstances, etc.

This is the difference IMO.

 

[ShrimpBrime]

I see so you do not enjoy helping someone out with a problem?
Discussing hardware in a civil manner without throwing insults.
Discussing trends in the industry with people from other markets
Learning about new software
Discussing reviews in a civilized manner.

It is those things that to me make TPU what it is. I would much rather have a discussion about what makes us PC users together than my CPU is better than yours or your GPU sucks because I think so. It is not like the PC that is the brain. Social media and forums are the mind of the internet. If not certain things would not have played out the way they did. Hypocrisy means noting today. Now we are adding AI to the mix.

I've been helping people for many years!! But I am not AI, so don't be mad at my point of view or feelings, I believe yours are just as valid.

AI is a tool. Like a gun. Like a horse. The horse was replaced by a car. The gun was replaced by a missile or bomb.

What if AI replaces humans!!! Speculation only. There is no evidence of AI taking over and ending it all.

If we use our tools properly, it can be a very safe thing to have.

What I've learned about CPUs is they are all the same. And these processors have AI integrated in them. And have for a long time. This bothers nobody seemingly.

 

[Bomby569]

Machines increased production but they didn't outright replace the human factor.

Artificial intelligence allied with advanced robotics has the potential to fully automate many jobs, replacing workers by high efficiency, 100% uptime machines that need no social security, no lunch or bathroom breaks, no time off, suffers from no health risks or unforeseen circumstances, etc.

This is the difference IMO.

you have no idea what you're talking about, there hasn't been a bigger disruption in history as the replacement of human labour in farms, agriculture. And there probably never will be anything like it again. AI won't replace all manual jobs, it can't.
See the scale of the numbers in China, for just a recent example.

https://www.researchgate.net/figure/Percent-of-the-Labor-Force-Employed-in-Agriculture-United-States-1800-to-2000-Sources_fig1_51117517

 

[ShrimpBrime]

If that's your example to make moot my point, then you clearly don't get it. Flipping someone the bird is NO WHERE NEAR the same thing as intentionally and falsely accusing another of something they did not do.

Yelling "Fire!" in a crowded theater when you know or believe there is a fire is perfectly legal. Yelling "Fire!" in a crowded theater when you know there is no fire is illegal. If you don't understand the difference then I feel sorry for you.


You don't know that. And that's the problem. Today, humans still control the actions of AI. But AI is evolving so quickly, it already is able act independently of humans. Fortunately, it has not done so (that we know of) maliciously. But tomorrow may be a different story.

How can AI replace the blue collar work force? That makes no sense. Blue collar jobs, by definition, are those which MUST be done manually - that is, physical tasks done by hand. AI is about reasoning, making decisions, or solving problems, not doing physical labor.

ROBOTICS may destroy the blue collar work force. And AI may direct the activities of those robotic machines. But AI will not replace physical labor jobs.

Yelling fire is freedom of speech. I did NOT say if that freedom of speech did or did not have a consequence. But you're still free to say fire regardless.

I understand AI could potentially create its self into something bad. But only if it has the freedom to do so.

 

[AusWolf]

you have no idea what you're talking about, there hasn't been a bigger disruption in history as the replacement of human labour in farms, agriculture. And there probably never will be anything like it again. AI won't replace all manual jobs, it can't.
See the scale of the numbers in China, for just a recent example.

I wish it replaced my manual job! But I agree, AI isn't a manual thing.

 

[the54thvoid]

Locked for editing or deleting.

Moderator intervention

One single warning!

The question of whether or not AI is a concern is highly relevant. What needs to be understood is the context of this discussion. As numerous members have stated, misinformation and political shenanigans have been around since 'Og' stole 'Ig's' counting sticks and conned the stone tribe in the 'who wants to fuck a neanderthal' referendum back in 60,000 BC.

You can easily replace 'AI' with 'guns', 'weaving machines', 'social media', or 'smart dildos.' Regardless, it's not the tech, but the (mis)use of it.

Besides the fact that TPUs forum rules restrict political pandering and toxicity, I don't even see why people need to mention politics or freedom of speech. AI, in and of itself, has nothing to do with that. And yes, I'm not blind to the irony that most people who wave the 'muh freedoms' flag don't actually understand what flag it is they're waving. Shrimpy said it - the power to give the (figurative) bird to authority. That's freedom of speech, but that's off-topic, regardless.

If I'm not treading on @lexluthermiester's toes, I'd suggest any further posts relate their concerns to the usage of AI - whether it nourishes or destabilises our human experience. Try commenting without pointy fingers or toxicity towards any group. Any group.

I'm available for PMs if you want to discuss why I've intervened. Please don't derail the thread by questionning them here (that's clearly against TPU forum rules.)

 

[lexluthermiester]

I'm going to echo 54thVoid's input. The idea was to have a civilized discussion about the feelings surrounding the use of AI.

It's understandable for everyone to have strong feelings. We just need to remember to take a step back and not go at each other personally and to avoid restricted topics.

AI is a scary thing for all the reasons we've discussed and more..

Let's face facts, this is a tough subject.

If I'm not treading on @lexluthermiester's toes

I'm good.

 

[Bill_Bright]

Thank you the54thvoid.

Really? All I've seen AI do so far is performing tasks, same as any other program.

Yes, really. I hope you are not suggesting because you have not seen something, that it cannot be?

You actually said 2 things to support my comment. First you said you have not seen "so far" AI do something. Part of my point was AI is evolving so fast, we humans are having difficulty keeping up. Then you gave an example of how AI can act independently. You tell it to write an essay, and it will independently go out, do the research and write the paper.

Yes, you told it the subject but in the future, it "may" be possible for AI, while doing that research for you, sees a problem and independently, tries to resolve it. I emphasize "may" because we don't know how it is likely to evolve - yet.

 

[R-T-B]

As a security expert, my concern level is very high. The cat is indeed out of the bag, but that certainly does not mean regulation is not something we should push before it gets WORSE.

I have zero confidence in regulation happening in the US however.

 

[freeagent]

AI for smart dildos and stuff like that would be fine, as a tool.. but not to run power grids, weapons systems, and important things like that..

 

[lexluthermiester]

As a security expert, my concern level is very high. The cat is indeed out of the bag, but that certainly does not mean regulation is not something we should push before it gets WORSE.

I have zero confidence in regulation happening in the US however.

Something needs to be done. If it's not strict regulation, then it's going to be worse than the wild west very soon. And your security concerns are yet another area of deep concern.

AI for smart dildos and stuff like that would be fine, as a tool..

Wait, what? How... nvrmnd.. I don't want to know..

 

[Bomby569]

What about self regulation, the cure some want to sell us, what do you guys think of it.
Can we have AI with clear boundaries working as intended, AI policing other AI, will it ever work?

Using AI to solve AI problems

 

[freeagent]

Something needs to be done. If it's not strict regulation, then it's going to be worse than the wild west very soon. And your security concerns are yet another area of deep concern.

I think the problem is money. There is too much to be made..

To me this is the new space race, how far can boundaries be pushed before we say oops?

 

[Macro Device]

AI for smart dildos

As an... enthusiast, let's call me that way, I STRONGLY disagree with this, unless you are an absolute masochist. AI's failures to recognise the situation can cause irrecoverable damage to your system. Dildos should be dumber than their users.

 

[lexluthermiester]

What about self regulation, the cure some want to sell us, what do you guys think of it.
Can we have AI with clear boundaries working as intended, AI policing other AI, will it ever work?

Using AI to solve AI problems

Self regulations of something this pervasive have never worked in the past. Rather dubious of it working at all.

To me this is the new space race, how far can boundaries be pushed before we say oops?

It could easily turn into the new nuke-arms race with the potential to be very harmful.

I'm going stand firm on the side of very strict regulations and harsh criminal penalties for abuse.

 

[Dr. Dro]

you have no idea what you're talking about, there hasn't been a bigger disruption in history as the replacement of human labour in farms, agriculture. And there probably never will be anything like it again. AI won't replace all manual jobs, it can't.
See the scale of the numbers in China, for just a recent example.

https://www.researchgate.net/figure/Percent-of-the-Labor-Force-Employed-in-Agriculture-United-States-1800-to-2000-Sources_fig1_51117517

We're skeptics on opposing sides of the coin, it's not that I don't understand but simply that we have a different take on it. The truth is that neither extreme is likely to happen; i.e. there will never be a full replacement of each and every worker but I do think a not insignificant percentage of the population will lose their livelihood to this technology.

We see signs of this in some businesses already, for example artists and voice actors who now are faced with competition in the form of readily accessible, affordable machine generated content. Even if it currently is of low overall quality, it only gets better over time.

Strikes like the SAG-AFTRA strike were motivated partially by the use of these technologies.

2023 SAG-AFTRA strike - Wikipedia

en.m.wikipedia.org

Anyway full disclaimer if I happen not to answer, I'm beat. Putting on my CPAP and catching some Z's.

 

[R-T-B]

What about self regulation, the cure some want to sell us, what do you guys think of it.
Can we have AI with clear boundaries working as intended, AI policing other AI, will it ever work?

Using AI to solve AI problems

AI is notoriously bad at detecting other AIs.

there will never be a full replacement of each and every worker

Honestly if every job was replaced it wouldn't even really be that bad. You could have universal income for everyone doing nothing save what they want and for once in history, society would still work.

That's cool. But the transition would be painful for sure.

 

[lexluthermiester]

AI is notoriously bad at detecting other AIs.

Maybe now, but it will not stay that way. Deep learning AI training can be molded into whatever task is at hand regardless of the type of task.

 

[freeagent]

Wait, what? How... nvrmnd.. I don't want to know..

Those are not my words, those are the words from the great Void above

 

[R-T-B]

Maybe not now, but it will not stay that way. Deep learning AI training can be molded into whatever task is at hand regardless of the type of task.

But on the counter, the AI will like get more sneaky at acting human on the same coin.

Hopefully its an arms race that we can win.

 

[kilo]

but not to run power grids, weapons systems, and important things like that..

Unfortunately, that's already happening with weapons systems.

 

[freeagent]

As an... enthusiast, let's call me that way, I STRONGLY disagree with this, unless you are an absolute masochist. AI's failures to recognise the situation can cause irrecoverable damage to your system. Dildos should be dumber than their users.

Ok, now I am intrigued..

Tell me more

 

[Bomby569]

AI is notoriously bad at detecting other AIs.

i meant behaviours, not AI's itself.

I'm no expert but i do see the appeal, if AI really becomes that good, why not use it to detect what other AI would be doing. If it ends by detecting bad behaviour from AI and humans alike that would not be a problem. Fight fire with fire

 

[R-T-B]

Unfortunately, that's already happening with weapons systems.

The US nuclear grid still uses offline 8" floppies. So not going there at least, lol.

 

[lexluthermiester]

Those are not my words, those are the words from the great Void above

I kinda glossed over that part..

The US nuclear grid still uses offline 8" floppies. So not going there at least, lol.

Not true, they upgraded years ago. But I digress..