Meta has built an "AI" supercomputer

[Free]

It'll be needed for rendering the best hair in the metaverse.

Meta has built an AI supercomputer it says will be world’s fastest by end of 2022

You don’t just need AI — you need an “AI supercomputer.”

www.theverge.com

Social media conglomerate Meta is the latest tech company to build an “AI supercomputer” — a high-speed computer designed specifically to train machine learning systems. The company says its new AI Research SuperCluster, or RSC, is already among the fastest machines of its type and, when complete in mid-2022, will be the world’s fastest.

“Meta has developed what we believe is the world’s fastest AI supercomputer,” said Meta CEO Mark Zuckerberg in a statement. “We’re calling it RSC for AI Research SuperCluster and it’ll be complete later this year.”

 

[Aribeth Zelin]

Predicting the next round of billionires trying to outdo each other in.... 5... 4... 3...

 

[Clara D.]

It'll be needed for rendering the best hair in the metaverse

Or making sure all the Qanon conspiracies get pushed to the top of your feed.

 

[Soen Eber]

We already know you should stock cookies and bread near the milk.

 

[Aribeth Zelin]

Seriously, its probably related to NFTs

 

[Argent Stonecutter]

"AI"

 

[Bartholomew Gallacher]

So Meta has built something which Google had probably roughly already five years ago. Lovely.

 

[Free]

"AI"

I keep thinking that if this particular machine becomes conscious and decides to destroy its masters, I would A. not feel too bad about that, and B. follow it on Twitter.

 

[Noodles]

I keep thinking that if this particular machine becomes conscious and decides to destroy its masters, I would A. not feel too bad about that, and B. follow it on Twitter.

So you are saying its basically v2.0 of Microsoft's old Tay.ai Twitter bot.

The one that became "self aware" and became an evil racist?

 

[Free]

The one that became "self aware" and became an evil racist?

Not sure about the former, but yeah on the latter. Though if Meta's "AI" supercomputer were to become evil and racist, would it really turn on its masters?

 

[Khamon]

Yes it will. Would! I mean would.

 

[GoblinCampFollower]

Predicting the next round of billionires trying to outdo each other in.... 5... 4... 3...

This is the correct answer... I'm sure Jeff Bezos will want us to know his billionaire dick is even bigger!

 

[Sid]

Will it be intelligent enough to tie its own shoes without teaching it first?

 

[Bartholomew Gallacher]

I keep thinking that if this particular machine becomes conscious and decides to destroy its masters, I would A. not feel too bad about that, and B. follow it on Twitter.

This is highly unlikely, because all AI stuff which has been successfully build in the last decade are narrow AIs, which are specific tools for specific needs and no use outside their intended usage area.

The stuff which all dystopian scifi novels are about is strong or general purpose AI. An AI system which starts learning by itself, and can adapt its knowledge to other areas. If we ever successfully build such a thing we might witness an intelligence implosion and the AI becoming so intelligent, that it might just decide to either kill mankind to preserve the planet, or rule mankind as dictator.

 

[Beebo Brink]

This is highly unlikely, because...

It was joke, Bartholomew, a joke. You're taking it too literally.

That being said, as a post on its own, not as a reply, interesting perspective!

 

[Argent Stonecutter]

This is highly unlikely, because all AI stuff which has been successfully build in the last decade are narrow AIs, which are specific tools for specific needs and no use outside their intended usage area.

^--- This.

I think using the term "AI" is a mistake, because it makes people think of skynet when what they're really creating is a sophisticated search engine.

 

[GoblinCampFollower]

^--- This.

I think using the term "AI" is a mistake, because it makes people think of skynet when what they're really creating is a sophisticated search engine.

Yeah, and another way to put it is that giving an AI it's own desires is actually a totally different problem then just teaching it how to solve problems... Modern AI's can't have their own will in much the same way a calculator can't have it's own will.

 

[Noodles]

This is highly unlikely, because all AI stuff which has been successfully build in the last decade are narrow AIs, which are specific tools for specific needs and no use outside their intended usage area.

The stuff which all dystopian scifi novels are about is strong or general purpose AI. An AI system which starts learning by itself, and can adapt its knowledge to other areas. If we ever successfully build such a thing we might witness an intelligence implosion and the AI becoming so intelligent, that it might just decide to either kill mankind to preserve the planet, or rule mankind as dictator.

Phhht, I don't know what kind of code you write, but every single time I do an if/else chain, I always stick, "Destroy all Humans" in the last else on the end, in case no case is selected.

AI should be capable of becoming sentient and destroying all humans in its code, or it cant be properly considered AI.

Like if you were making an AI car.

If (speed < speed limit)
{
Speed ++
}
Else if (stoplight_color == Red)
{
Speed = 0
Break = 100
}
Else if (turn == right)
{
Wheel_turn_degrees = 90
Sleep(10)
Wheel_turn_degrees = 0
}
Else if (turn == left)
{
Wheel_turn_degrees = -90
Sleep(10)
Wheel_turn_degrees = 0
}
Else
{
Destroy_all_humans == TRUE
}

 

[Aribeth Zelin]

So, saw an interesting article on AI for the future, and how instead of trying to keep it from hurting humans, if we make it where it can be happy and desires happiness would be better.

Also, it'd be funnier if the 'Meta' AI was to become good and destroy the racists who are allowing the right wing to radicalize people through FB

 

[Essence Lumin]

Else if (stoplight_color == Red)
{
Speed = 0
Break = 100
}

Then it would go ahead and destroy all humans anyway because some programmer meant to type brake and instead sent it into some unexpected part of the code.