Economy Sam Altman "OpenAI founder" is seeking 5 trillion dollars to build AI chip technology.

[PEB]

"Reports say that Sam Altman plans to raise $5 to $7 trillion to build a network of fabs to produce enough chips for artificial intelligence (AI) applications, raising many eyebrows in the industry. Nvidia's Jensen Huang said that the architectural innovation of AI processors is more important than the quantity of these processors. Now, Jim Keller, a legendary CPU developer who now works at Tenstorrent developing AI and HPC processors, essentially claims the same thing."

Love Jim likely the greatest chip designer in history. Got his start in my backyard at Digital Equipment now head architect an making a pretty good joke.

 

[DC23]

Data centres the size of cities soon enough. Guess the US has the space tbf.

 

[Ruprecht]

It is an INSANE amount of money. But, the tech needed to do what I think they are aiming for (a completely true AI) is going to need tech well beyond what we have now.

Of course AI processing needs developing, and the barriers to entry for chip design and production are enormous.
The point is Sam's effectively asking to be made the largest tech company by investors, setting up a rival to Nvidia with vertical integration, purely on the speculation about what general AI can achieve. This just months after the embarrassing collapse of their corporate structure.
In terms of demonstrating that use, all they have is the chat bots, image and video generation. I think that's probably not enough for investors to make that leap, regardless of the current buzz around AI.
I have chatgpt integrated at work for managed services, and it hasn't really proven it's worth. Aside from the comical inability to meaningfully paraphrase or respond to Australian slang, colloquialisms and humour in a corporate context. Script generation hasn't really been the time saver you might expect either, due to the lack of reliability. Even when it's essentially just plugging the relevant variables into an existing template.

 

[RoastBeast]

Have Russia attack him and Biden will hand over 5 trillion.

 

[Michaelangelo]

Oh, the progress is amazing. Channels are already comparing the leap in the Sora demos vs that year old AI video of Will Smith eating spaghetti.



No comparison.

5 trillion USD is a lot of money though. Apple is the highest valuated company at the moment at @2.82 trillion. The GDP of my country is @1.7 trillion.

Actually Microsoft surpassed Apple

Microsoft also owns a majority of OpenAI

Wouldn’t be surprised for them to hit 5 trillion in the next few years

 

[Super_Nintendo]

View attachment 1029905

can't watch it in stereo. mono track or nothing.

 

[Ruprecht]

Actually Microsoft surpassed Apple

Microsoft also owns a majority of OpenAI

Wouldn’t be surprised for them to hit 5 trillion in the next few years

Looks like it's because Apple dropped off. Guess the Vision Pro hasn't maintained the hype.
Microsoft owns 49% of OpenAI, it's still a minority.
Microsoft's current trajectory might have them hitting 5 trillion soon, but their cloud growth has slowed right down, so perhaps not. That said Azure AD is still pretty dominant in the industry and offers the best opportunity for AI integration.

 

[High Test With da Best]

"Reports say that Sam Altman plans to raise $5 to $7 trillion to build a network of fabs to produce enough chips for artificial intelligence (AI) applications, raising many eyebrows in the industry. Nvidia's Jensen Huang said that the architectural innovation of AI processors is more important than the quantity of these processors. Now, Jim Keller, a legendary CPU developer who now works at Tenstorrent developing AI and HPC processors, essentially claims the same thing."

Love Jim likely the greatest chip designer in history. Got his start in my backyard at Digital Equipment now head architect an making a pretty good joke.

"...making a pretty good joke"

What's that mean?

 

[Ruprecht]

"Reports say that Sam Altman plans to raise $5 to $7 trillion to build a network of fabs to produce enough chips for artificial intelligence (AI) applications, raising many eyebrows in the industry. Nvidia's Jensen Huang said that the architectural innovation of AI processors is more important than the quantity of these processors. Now, Jim Keller, a legendary CPU developer who now works at Tenstorrent developing AI and HPC processors, essentially claims the same thing."

Love Jim likely the greatest chip designer in history. Got his start in my backyard at Digital Equipment now head architect an making a pretty good joke.

Does Digital Equipment still exist, I thought they disappeared in the late '80s?

Spoiler: Nerds

 

[Dirt Road Soldier]

Of course AI processing needs developing, and the barriers to entry for chip design and production are enormous.
The point is Sam's effectively asking to be made the largest tech company by investors, setting up a rival to Nvidia with vertical integration, purely on the speculation about what general AI can achieve. This just months after the embarrassing collapse of their corporate structure.
In terms of demonstrating that use, all they have is the chat bots, image and video generation. I think that's probably not enough for investors to make that leap, regardless of the current buzz around AI.
I have chatgpt integrated at work for managed services, and it hasn't really proven it's worth. Aside from the comical inability to meaningfully paraphrase or respond to Australian slang, colloquialisms and humour in a corporate context. Script generation hasn't really been the time saver you might expect either, due to the lack of reliability. Even when it's essentially just plugging the relevant variables into an existing template.

I think he has ever right to ask that, though. A few years ago every laughed at AI generated images, saying it will never do this or that. But now just a few scant years, or even months later? And we have shit coming out that everyone thought would take years if not decades. The speed at which this tech is advancing is insane, and needs huge investments to keep up at this pace. Frankly, the tech is far outpacing the hardware needed.

Now, some could argue we SHOULDN'T be going down this route, for a number of reasons. Some of which even my redneck ass can agree with. But at the same time, advancement in scientific knowledge waits for nobody, so someone has to be at the front. And at least Sam is a known quantity, if a bit of a loon.

 

[Ruprecht]

I think he has ever right to ask that, though. A few years ago every laughed at AI generated images, saying it will never do this or that. But now just a few scant years, or even months later? And we have shit coming out that everyone thought would take years if not decades. The speed at which this tech is advancing is insane, and needs huge investments to keep up at this pace. Frankly, the tech is far outpacing the hardware needed.

Now, some could argue we SHOULDN'T be going down this route, for a number of reasons. Some of which even my redneck ass can agree with. But at the same time, advancement in scientific knowledge waits for nobody, so someone has to be at the front. And at least Sam is a known quantity, if a bit of a loon.

He's not just talking about research and design though, that 5-7 trillion is for complete vertical integration. OpenAI would own the land, own the fabs, own the datacentres, and develop, own and control the AI being deployed and all the hardware/infrastructure.
I'm in no way any sort of neo-luddite or doom sayer, but the idea of setting them up in that sort of position of monopolistic power, especially with their recent corporate failure and the shift from their original stated values and priorities, seems ridiculous. Especially before we've even seen the commercial viability of any product.
I mean dream big by all means, but if that's not just hype and marketing it's actually a bit worrying. Not just because of the sum of money, but the nature of the proposition and the fact Altman's supposedly approached the US government for approval which will exempt him from existing antitrust legislation.

 

[Whippy McGee]

They will want to build some enormous underground absolutely frozen beyond normal space with ultra dry air to run whatever they have in mind. The heat produced would be crazy. They also need to have a small to micro nuclear reactor to power it.... but $5 Trillion... GTFO damn grifter.

 

[Michaelangelo]

Looks like it's because Apple dropped off. Guess the Vision Pro hasn't maintained the hype.
Microsoft owns 49% of OpenAI, it's still a minority.
Microsoft's current trajectory might have them hitting 5 trillion soon, but their cloud growth has slowed right down, so perhaps not. That said Azure AD is still pretty dominant in the industry and offers the best opportunity for AI integration.

Yeah should have said significant stake in OpenAI instead

These valuations are getting absurd

You have corporations with enough wealth to become nations unto themselves

 

[N13]

Get that mang a quantum computer

 

[moreorless87]

AI chips which know what dip you need?

 

[hitcher]

Yeah should have said significant stake in OpenAI instead

These valuations are getting absurd

You have corporations with enough wealth to become nations unto themselves

the wealth is based on these asinine valuations. If the companies start selling their own stock in relevant number stock would tank. Hence the wealth is just theoretical in nature.

 

[HeLLMuTT]

 

[Osculater]

The best I can do Sam is aboot tree fiddy

 

[Bunk Moreland]

Has he tried committing a massive amount of fraud then asking people to donate to raise a ridiculous amount of money? That usually works here in the USA.

 

[Dirt Road Soldier]

He's not just talking about research and design though, that 5-7 trillion is for complete vertical integration. OpenAI would own the land, own the fabs, own the datacentres, and develop, own and control the AI being deployed and all the hardware/infrastructure.
I'm in no way any sort of neo-luddite or doom sayer, but the idea of setting them up in that sort of position of monopolistic power, especially with their recent corporate failure and the shift from their original stated values and priorities, seems ridiculous. Especially before we've even seen the commercial viability of any product.
I mean dream big by all means, but if that's not just hype and marketing it's actually a bit worrying. Not just because of the sum of money, but the nature of the proposition and the fact Altman's supposedly approached the US government for approval which will exempt him from existing antitrust legislation.

Guess what. The first country that gives them the rules, will own the right to MAKE the rules. So if a coalition of whatever countries sinks the money into this tech, well I guess we all get to suffer from what THEY want. And if it happens to be a giant bunch of {inserted left wing tard words} or {inserted right wing tard words} well I guess they win.

This isnt a game. One way or the other, someone is going to win and there will be bloodshed. Because the world is by no means ready for this shit, but its happening and we are all just along for the ride.