The AI apocalypse

[Hondaracer]

So I looked around a bit for an AI image generator and eventually downloaded this Leonardo.AI app, fuck this shot is pretty crazy.. this app has a tonne of features and settings and I’m getting pretty crazy results with very basic input,

“Cat in a cowboy outfit shooting his pistols”





Hulk hogan wining the masters at Augusta national in front of a large crowd



Don cherry kissing Vladimir Putin on the back of a yacht in the sunset



Was trying to think of the most random shit to generate lol

What’s VERY impressive/scary is when chatGTP first came out my wife and I were playing around with its image generator and it was coming up with some very basic/rudementary images like barely recognizable.

Now in a matter of what, 8 months, you’re getting results like that with very basic queries? Pretty crazy

[RabidRat]

Yeah I'm really getting kinda paranoid about posting personal stuff online now. Seems relatively easy for that to be scraped and faked.

You can still kinda tell those images are CGI, but it's getting really close. And with such little effort!

[sonick]

CEO of OpenAI abruptly fired, reasons under speculation but seemed pretty sudden and surprising in the industry

https://www.theverge.com/2023/11/17/...m-altman-fired

Probably something crazy happening behind the scenes to be fired so suddenly with very little smoke.

[Manic!]

And look who was just hired by Microsoft.

https://www.theverge.com/2023/11/20/...ployees-openai

Microsoft hires former OpenAI CEO Sam Altman / Greg Brockman, OpenAI co-founder, is also joining Microsoft to lead a new advanced AI research team

https://twitter.com/satyanadella/sta...ployees-openai

[Manic!]

https://www.theverge.com/2023/11/20/...oft-sam-altman

505 of the 700 employees at openai threaten to leave ans join Microsoft if Altman is not reinstated and the board does not resign.

[Traum]

I dunno about you guys, but I'm getting a kick out of the Board of Directors getting burned so bad by the backfiring of their plans lol~

[sonick]

Yeah I am looking forward to hearing the inner workings of how all this came to be.

Satya was probably licking his chops to snag up Sam when he heard OpenAI fired him. What a gangster move.

[Manic!]

https://www.datacenterdynamics.com/e...-chip-venture/

Looks like Sam may have been trying to raise billions for a new AI chip company to compete with NVIDIA at the same time the board was trying to raise money.

[Traum]

Personally, I am not sure why Altman wants to pursue having its own AI chip company. It is true that nVidia dominates the market at the moment, but with OpenAI's partnership with Microsoft, it seems far less risky and cheaper to me to just go along with Microsoft's design instead of trying to do your own. At the end of the day, the bottleneck with chip manufacturing really comes from TSMC's manufacturing capacity, since they are pretty much the only company that can reliably, competitively, and consistently produce the latest generation of high end chips.

But of course, Altma is the CEO, and I'm just a bum shooting my mouth off on the interwebs~

[Manic!]

He's back.

https://twitter.com/OpenAI/status/17...077370115?s=20

[Manic!]

Originally Posted by Traum

Personally, I am not sure why Altman wants to pursue having its own AI chip company. It is true that nVidia dominates the market at the moment, but with OpenAI's partnership with Microsoft, it seems far less risky and cheaper to me to just go along with Microsoft's design instead of trying to do your own. At the end of the day, the bottleneck with chip manufacturing really comes from TSMC's manufacturing capacity, since they are pretty much the only company that can reliably, competitively, and consistently produce the latest generation of high end chips.

But of course, Altma is the CEO, and I'm just a bum shooting my mouth off on the interwebs~

Money. Why use Microsoft's or NVIDIA's chip when you can raise billions in the middle east.

[Manic!]

https://www.reuters.com/technology/s...gh-2023-11-22/

Nov 22 (Reuters) - Ahead of OpenAI CEO Sam Altman’s four days in exile, several staff researchers wrote a letter to the board of directors warning of a powerful artificial intelligence discovery that they said could threaten humanity, two people familiar with the matter told Reuters.

The previously unreported letter and AI algorithm were key developments before the board's ouster of Altman, the poster child of generative AI, the two sources said. Prior to his triumphant return late Tuesday, more than 700 employees had threatened to quit and join backer Microsoft (MSFT.O) in solidarity with their fired leader.

[SkinnyPupp]

So it's learning like a human now, instead of being a glorified search engine

Interesting

[underscore]

Welp, time to buy a black trench coat and sunglasses I guess

[mikemhg]

You guys have to keep in mind some of these people on the board are part of EA (Effective Altruism) Movement, which is almost a quasi evangelical religious ideology.

Some of these guys have a god-complex of themselves, which is quite ridiculous.

AI won't take over the world and destroy us, at least not anytime soon. We're nowhere close to such technology.

Much of this is simple dick jerking to pump up excitement and investment into their platform and products.

This shit drives me crazy.

[underscore]

...can I still wear the trenchcoat?

[murd0c]

Originally Posted by SkinnyPupp

So it's learning like a human now, instead of being a glorified search engine

Interesting

This really makes one think, if this is now happening just think about how quick it will grow and basically turn into a sentient being. That being said the whole simulation theory conversation comes into play how who says the same thing hasn't happened to us?

[mikemhg]

Originally Posted by murd0c

This really makes one think, if this is now happening just think about how quick it will grow and basically turn into a sentient being. That being said the whole simulation theory conversation comes into play how who says the same thing hasn't happened to us?

We may never reach machine sentience, these things are nothing but elaborate algorithms using immense computing power to brute force process the calculations. That's not sentience as much as we try to pretend it is.

We're no where close to sentience, we may never even see this in our lifetimes.

[SkinnyPupp]

It doesn't need to reach sentience for AI to be paradigm shifting... And the way things are going, we need a paradigm shift for there to be anything but utter catastrophe in our future (and I don't mean our grandkids' future, I mean ours).

[mikemhg]

You're wrong, Skinny.

AI will be paradigm shifting, in fact it is already. From process automation, media, journalism, administrative services, health care research, all of which will and are being shifted by AI. It will affect labor markets, and the way we engage multitudes of work, most definitely.

What we need to stop pretending is that AI will be Skynet, some amorphous type of thing that will think for itself and destroy the world.

That type of talk is utterly ridiculous.

[SkinnyPupp]

Originally Posted by mikemhg

You're wrong, Skinny.

AI will be paradigm shifting, in fact it is already. From process automation, media, journalism, administrative services, health care research, all of which will and are being shifted by AI. It will affect labor markets, and the way we engage multitudes of work, most definitely.

What we need to stop pretending is that AI will be Skynet, some amorphous type of thing that will think for itself and destroy the world.

That type of talk is utterly ridiculous.

What am I wrong about? It looks like we're agreeing...

The paradigm that has to shift is how currency works. We (humanity) need to change that to be able to progress, or things will continue towards catastrophe (climate catastrophe notwithstanding). For AI to be usefully paradigm shifting, it needs to do something like that. Get the wealth out of the hands of "the 1%" (ancient term I know but it fits)

I'm not saying I am hopeful this will happen or that it's even possible, but other than discovering infinite free energy, I don't see what else can happen. If it's not this, or aliens visiting to fix everything, maybe it's AI

[underscore]

Originally Posted by mikemhg

From process automation, media, journalism, administrative services, health care research, all of which will and are being shifted by AI.

What part of it? Because at least the stuff I work in hasn't changed at all yet, and largely runs on hardware a decade or two out of date lol.

Originally Posted by mikemhg

What we need to stop pretending is that AI will be Skynet, some amorphous type of thing that will think for itself and destroy the world.

I'm more concerned about too powerful an AI being controlled by humans than of it thinking for itself. Because humans will use it to influence the far too large group of people who don't understand how it works.

[Hehe]

The holy-grail of AI is a system where it's capable of learning and expressing its own opinions and thoughts *AND* acting on it.

Currently, whether it's ChatGPT, Diffusion or whatever AI out there, they are confined within the first 2 parts. They cannot ACT by itself unless asked for. Say you want to use Diffusion to create some pictures, you need to ask for it. Then they would go through the first 2 steps and give you the final step.

But what if there's an AI that's no longer bound by such limitation? It learns whatever it can find, it gives whatever opinion it has and formulate whatever conclusion it has from the first 2 steps?

Yes, we might be able to hardcode cultural, moral, discipline... etc into the model. But the question becomes what should be "right", and what should be "wrong"? Many things in the world is not just black and white.

Human can barely make that distinction and we have many many arguments as often right and wrong is only a line that separates them. And if we cannot develop to a level where we achieve universal agreement, what are we supposed to teach the AI?

An AI without limits is scary, to say the least. Especially in an era when much of what we call "assets" are just numbers in a system. If an AI can think by itself, and has a nearly unlimited processing power, I don't think there are many things that stand in their way should it decide to go rogue. Worse, it doesn't even realize that that's bad. It thinks that those securities put in place are *wrong* as they prevent the AI from doing "the right thing".

We don't need AI to become Skynet to be worry about it. Knowledge is power. If a person who has control over this super AI decides that the AI can help him to achieve world domination... what's in his/her/their way? Computers with enough processing power can see the potential outcomes of any decision a million miles away and make the choice of the best way to achieve what their master wants to do.

We as humans cannot access the amount of data that AI can and have no where near the processing power. If used incorrectly, AI can change the world as we know it.

When Homo Sapiens appeared, it was followed by the extinction of Neanderthals. We are the Neanderthals in this case. Was it better for Earth? Yeah.. maybe.. We as human made a leap forward. But are you ok with a world where only AI exists?

To reflect on Skinny's idea... "Get the wealth out of the 1%". What if it thinks that as you take away from 1%, someone from the 99% remaining would become the new 1%? Thus the best way is just find a way to eliminate human. It's perfection. No one needs to suffer anymore or fight for wealth.

[mikemhg]

Originally Posted by underscore

What part of it? Because at least the stuff I work in hasn't changed at all yet, and largely runs on hardware a decade or two out of date lol.



I'm more concerned about too powerful an AI being controlled by humans than of it thinking for itself. Because humans will use it to influence the far too large group of people who don't understand how it works.

You'll see it slowly creep in, if you use Office or Google suites it's already there in small ways. Microsoft most notably has been creeping it into their products. For example the Viva emails you might get through your Teams/Outlook are using AI to track stats/engagement for reporting purposes, etc.

Any search engine you utilize is already incorporating AI.

It's hard to introduce AI to legacy platforms as it currently stands without a lot of work, so if your company is using old systems, you likely won't see much of a change until they invest in newer tech (which companies are doing).

[underscore]

Originally Posted by mikemhg

You'll see it slowly creep in, if you use Office or Google suites it's already there in small ways. Microsoft most notably has been creeping it into their products. For example the Viva emails you might get through your Teams/Outlook are using AI to track stats/engagement for reporting purposes, etc.

Any search engine you utilize is already incorporating AI.

It's hard to introduce AI to legacy platforms as it currently stands without a lot of work, so if your company is using old systems, you likely won't see much of a change until they invest in newer tech (which companies are doing).

Into the general Office type stuff sure, but not into actual automation.

I have enough dumbass customers telling me the code changed itself already, no way is anyone gonna let it actually do that