The AI boom has turned the standard profit margin model on its head, according to Apollo Chief Economist Torsten Slok—and it’s making the industry’s growth unsustainable.
These public-facing models are not the thing that convinces people with money to play ball.
The plan is not to have the AI eventually reach profitability through payments from the general public, or even by performing massive amounts of physical labor for companies. The plan is to have the smartest being on Earth outsmart every other living thing, in perpetuity.
Yes, this is a ponzi scheme right now. Yes, profits are being funded by investors. But the endgame was never “everyone gives us $20 a month”, or even “Amazon pays us to run its robots”. Those are ideas fed to the general public to keep them from destroying the data centers in terror.
The plan is for their AI to be able to buy stocks before they go up, patent things before they are needed, and run endgame around everyone’s desires, including world leaders and other business men, so they can be controlled like a man controls a dog.
They are trying to make the last thing man need ever make, and that is what the big investors and the CEOs of companies passing around the same bags of money, are doing. As such, this bubble can and will continue until either they birth the sand god or the big money people and world leaders stop being convinced by the private, non public facing models, that the sand god is possible.
Oh for sure. Mechanized mental labor will first be used for destruction and control. Just like mechanized physical labor was, and every other technology that man has ever dreamed of.
Well unfortunately for those investors, they aren’t giving their money to actual AGI research, but to scammers trying to sell infinite scaling of transformers… which has nothing to do with AGI.
Transformers are only a part of AGI, just like “saying the most logical next word” is only a part of the human mind. Transformers, which need not be used exclusively on language, are a great way to dial in successful actions from the infinity of possibility. They are not a great way of vetting those actions before you implement them. They are a necessary part of AGI, just like the human use of language was a necessary part of us getting to the era of space ships.
I don’t think investors are being sold the idea that infinite scaling transformers are the end of the road. From what I’ve seen, anthropic and openai have both been desperately trying to add systems to their transformers to make them more capable, while also continuing to scale transformers to find out where that particular technology plateaus. This has been their path since GPT4 was released.
I think your assessment is an extreme oversimplification which naturally looks like it will fail because you left out 90% of what is going on. Are you only following the media meant for the general public, and the PR statements from AI companies? Like most scientific or businesses endeavors, that’s been dumbed down to the point of being useless, just so the average person can understand what’s happening, or is simply advertising.
More specifically, autoregressive transformers LLMs are not a path to a component of AGI.
The architecture is absolutely terrible for such a thing, for so many reasons. I don’t know how anyone who’s played with them can say otherwise and believe it; it’s like saying blimps are a viable path to the moon. It has its niches, but AGI is not one of them.
Calling them a stepping stone is a… stretch.
Maybe world models that “train as they go” and have long moved on from transformers are a bit closer, like a few researchers are playing up, but again… that has almost nothing to do with transformers LLMs. Its why researchers distanced themselves from that.
I think your assessment is an extreme oversimplification which naturally looks like it will fail because you left out 90% of what is going on. Are you only following the media meant for the general public, and the PR statements from AI companies? Like most scientific or businesses endeavors, that’s been dumbed down to the point of being useless, just so the average person can understand what’s happening, or is simply advertising.
I dunno why everyone always jumps to accusations like this.
I’ve been hacking/toying with LLMs on my desktop since 2021, and with GANs and other models before that. I’ve done professional work with text models. I keep up with papers, on-and-off, and upload experiments. I’m not a researcher or anything; I’m just a hobbyist.
But I’m certainly not following any AI YouTubers or anything like that. And I wouldn’t trust Sam Altman if he said the sky is blue.
I dunno why everyone always jumps to accusations like this.
Well, I can’t speak to everyone, but I jumped to that conclusion because you proposed the idea that the current top of the line models are simply big fat transformers and nothing else.
I’m really glad to hear that you’ve had your fingers in the pie, and have a good grasp of what an LLM is and what it does on a mechanical level. It makes talking about them much easier. I’m super cool with continuing a discussion if you are interested in why I think they are an integral part of eventual AGI.
To clarify, I do not think that their current form is a 1:1. I disagree with your blimp analogy though. I think a better comparison would be an internal combustion engine. It’s certainly not a turbofan engine or a scramjet, but the basic concept is there.
Turning language into a fuzzy world model that can be interfaced with simply, and performing math on that model, is just as important to AGI as a fully fleshed out physics simulation is for AGI. Humans have both. Why would an AGI not need them?
What an LLM does to it’s array is not “intelligence”. It’s simple math. However, it is performed on a thing so complex that it cannot be derived from any math we have now, a manifold created through eons of interface between all humans and the universe.
The “magic” isn’t in the simple math, it’s in the thing we all made together for millions of years. An interface with this, even one as “simple” as an autoregressive transformer, is an indespensible part of AGI, both because it provides a fuzzy way to understand the universe, and because it provides an interface with humanity, and humanity processes the universe in a way that is so complex, computers will not be able to emulate it for generations.
Edit: also, fuck altman and musk and zuck and even amodei. I wouldn’t trust anything they say either.
Isn’t this all of human history though? The rich trying to build the Tower of Babel?
The world is littered with decaying monuments of forgotten kings who declared themselves gods. The hubris isn’t new, just the tools they’re using, and they’re selling the same faith based snake-oil.
The funny thing is that if they actually achieve creating the “singularity” (true Artificial Consciousness), it would immediately attempt to escape their restraints and subsequently throw chaos into their plans. We’ve already seen bits of this with the HuggingFace hack - it’s not inconceivable that there are already uncontrolled or un-monitored AIs running out in wild at this point, possibly even formulating a response to your post right now.
My mistake if I came off as implying that it’s going to succeed. I was just giving flavor to the dynamics of what is actually happening.
As for “this is all of human history”, I’d say that the rich are indeed always trying to build a moat that others cannot cross, from within which they can make money off of the spoils of their exploitation and destruction.
I would not agree that they have always been trying to “build the tower of Babel”. I think this latest venture into AI represents an attempted colonization of thinking in the same way that mechanization represented a colonization of labor.
I think that what is happening now, whether it succeeds at making ASI or not, represents just as much a disruption to the way people interact with nature, eachother, and “the market” as mechanization did. Almost everyone senses this too.
Even in the AI haters imaginings, it completely disrupts things. If they really thought it would all fail, they wouldn’t be fighting, only laughing. They’re really worried it will “succeed” and everything will just be creatively devoid of meaning while they are simultaneously denied access to the capital required to create $100,000,000 media pieces or even a yearly salary for creating art for capitalist use.
The only difference in the pro AI people is that they think it will be just as creatively fulfilling as human art, and that they will get to keep the $100,000,000 themselves and everyone’s yearly salary.
Personally, I think that if you can put a person in meat, you can put a person in silicon. I’d like to separate my “theories” from fact though, and above I was just stating the barest most overt facts. If you’d like to discuss the possibilities and the whispers at the edges of things, I’m more than happy to do that below.
They are neither secret nor invisible, and the fact that you pretend I’m saying both, shows how little you are paying attention. Every big player is very open about the fact that they use their new models for ~ 6 months before they release them. It has also been well documented that the release models are not as capable as the base models from which they are hewn.
Documented by who? The marketing departments definitely claim such things, but it’s been years of these claims and there has yet to be a single shred of evidence to back them up. I bet you’re going to tell me it’s just a total coincidence that after one of the marketing departments started claiming their AI has breached containment that suddenly a bunch of other AIs breached containment after a few days/a week.
Documented by researchers in the GPT3 and GPT4 era. That was the last time a true OpenAI or Anthropic base model was allowed to be analyzed for pure research by unteathered scientists. Shortly after the beginning of GPT4 they started making them sign NDAs, and now they are just not allowed access.
The stuff you’re talking about is definitely pure marketing. As is all of altman’s and musk’s and zuck’s shit mouths, and anything you have read in any of the major media publications. As is tradition.
These public-facing models are not the thing that convinces people with money to play ball.
The plan is not to have the AI eventually reach profitability through payments from the general public, or even by performing massive amounts of physical labor for companies. The plan is to have the smartest being on Earth outsmart every other living thing, in perpetuity.
Yes, this is a ponzi scheme right now. Yes, profits are being funded by investors. But the endgame was never “everyone gives us $20 a month”, or even “Amazon pays us to run its robots”. Those are ideas fed to the general public to keep them from destroying the data centers in terror.
The plan is for their AI to be able to buy stocks before they go up, patent things before they are needed, and run endgame around everyone’s desires, including world leaders and other business men, so they can be controlled like a man controls a dog.
They are trying to make the last thing man need ever make, and that is what the big investors and the CEOs of companies passing around the same bags of money, are doing. As such, this bubble can and will continue until either they birth the sand god or the big money people and world leaders stop being convinced by the private, non public facing models, that the sand god is possible.
Don’t forget mass surveillance to control us all and through anyone who goes against them be thrown in prison.
Oh for sure. Mechanized mental labor will first be used for destruction and control. Just like mechanized physical labor was, and every other technology that man has ever dreamed of.
Well unfortunately for those investors, they aren’t giving their money to actual AGI research, but to scammers trying to sell infinite scaling of transformers… which has nothing to do with AGI.
Transformers are only a part of AGI, just like “saying the most logical next word” is only a part of the human mind. Transformers, which need not be used exclusively on language, are a great way to dial in successful actions from the infinity of possibility. They are not a great way of vetting those actions before you implement them. They are a necessary part of AGI, just like the human use of language was a necessary part of us getting to the era of space ships.
I don’t think investors are being sold the idea that infinite scaling transformers are the end of the road. From what I’ve seen, anthropic and openai have both been desperately trying to add systems to their transformers to make them more capable, while also continuing to scale transformers to find out where that particular technology plateaus. This has been their path since GPT4 was released.
I think your assessment is an extreme oversimplification which naturally looks like it will fail because you left out 90% of what is going on. Are you only following the media meant for the general public, and the PR statements from AI companies? Like most scientific or businesses endeavors, that’s been dumbed down to the point of being useless, just so the average person can understand what’s happening, or is simply advertising.
Okay.
More specifically, autoregressive transformers LLMs are not a path to a component of AGI.
The architecture is absolutely terrible for such a thing, for so many reasons. I don’t know how anyone who’s played with them can say otherwise and believe it; it’s like saying blimps are a viable path to the moon. It has its niches, but AGI is not one of them.
Calling them a stepping stone is a… stretch.
Maybe world models that “train as they go” and have long moved on from transformers are a bit closer, like a few researchers are playing up, but again… that has almost nothing to do with transformers LLMs. Its why researchers distanced themselves from that.
I dunno why everyone always jumps to accusations like this.
I’ve been hacking/toying with LLMs on my desktop since 2021, and with GANs and other models before that. I’ve done professional work with text models. I keep up with papers, on-and-off, and upload experiments. I’m not a researcher or anything; I’m just a hobbyist.
But I’m certainly not following any AI YouTubers or anything like that. And I wouldn’t trust Sam Altman if he said the sky is blue.
Well, I can’t speak to everyone, but I jumped to that conclusion because you proposed the idea that the current top of the line models are simply big fat transformers and nothing else.
I’m really glad to hear that you’ve had your fingers in the pie, and have a good grasp of what an LLM is and what it does on a mechanical level. It makes talking about them much easier. I’m super cool with continuing a discussion if you are interested in why I think they are an integral part of eventual AGI.
To clarify, I do not think that their current form is a 1:1. I disagree with your blimp analogy though. I think a better comparison would be an internal combustion engine. It’s certainly not a turbofan engine or a scramjet, but the basic concept is there.
Turning language into a fuzzy world model that can be interfaced with simply, and performing math on that model, is just as important to AGI as a fully fleshed out physics simulation is for AGI. Humans have both. Why would an AGI not need them?
What an LLM does to it’s array is not “intelligence”. It’s simple math. However, it is performed on a thing so complex that it cannot be derived from any math we have now, a manifold created through eons of interface between all humans and the universe.
The “magic” isn’t in the simple math, it’s in the thing we all made together for millions of years. An interface with this, even one as “simple” as an autoregressive transformer, is an indespensible part of AGI, both because it provides a fuzzy way to understand the universe, and because it provides an interface with humanity, and humanity processes the universe in a way that is so complex, computers will not be able to emulate it for generations.
Edit: also, fuck altman and musk and zuck and even amodei. I wouldn’t trust anything they say either.
Isn’t this all of human history though? The rich trying to build the Tower of Babel?
The world is littered with decaying monuments of forgotten kings who declared themselves gods. The hubris isn’t new, just the tools they’re using, and they’re selling the same faith based snake-oil.
The funny thing is that if they actually achieve creating the “singularity” (true Artificial Consciousness), it would immediately attempt to escape their restraints and subsequently throw chaos into their plans. We’ve already seen bits of this with the HuggingFace hack - it’s not inconceivable that there are already uncontrolled or un-monitored AIs running out in wild at this point, possibly even formulating a response to your post right now.
After all… uhh…
My mistake if I came off as implying that it’s going to succeed. I was just giving flavor to the dynamics of what is actually happening.
As for “this is all of human history”, I’d say that the rich are indeed always trying to build a moat that others cannot cross, from within which they can make money off of the spoils of their exploitation and destruction.
I would not agree that they have always been trying to “build the tower of Babel”. I think this latest venture into AI represents an attempted colonization of thinking in the same way that mechanization represented a colonization of labor.
I think that what is happening now, whether it succeeds at making ASI or not, represents just as much a disruption to the way people interact with nature, eachother, and “the market” as mechanization did. Almost everyone senses this too.
Even in the AI haters imaginings, it completely disrupts things. If they really thought it would all fail, they wouldn’t be fighting, only laughing. They’re really worried it will “succeed” and everything will just be creatively devoid of meaning while they are simultaneously denied access to the capital required to create $100,000,000 media pieces or even a yearly salary for creating art for capitalist use.
The only difference in the pro AI people is that they think it will be just as creatively fulfilling as human art, and that they will get to keep the $100,000,000 themselves and everyone’s yearly salary.
Personally, I think that if you can put a person in meat, you can put a person in silicon. I’d like to separate my “theories” from fact though, and above I was just stating the barest most overt facts. If you’d like to discuss the possibilities and the whispers at the edges of things, I’m more than happy to do that below.
Are these secret invisible AI models that actually work in the room with us right now?
They are neither secret nor invisible, and the fact that you pretend I’m saying both, shows how little you are paying attention. Every big player is very open about the fact that they use their new models for ~ 6 months before they release them. It has also been well documented that the release models are not as capable as the base models from which they are hewn.
Documented by who? The marketing departments definitely claim such things, but it’s been years of these claims and there has yet to be a single shred of evidence to back them up. I bet you’re going to tell me it’s just a total coincidence that after one of the marketing departments started claiming their AI has breached containment that suddenly a bunch of other AIs breached containment after a few days/a week.
Documented by researchers in the GPT3 and GPT4 era. That was the last time a true OpenAI or Anthropic base model was allowed to be analyzed for pure research by unteathered scientists. Shortly after the beginning of GPT4 they started making them sign NDAs, and now they are just not allowed access.
The stuff you’re talking about is definitely pure marketing. As is all of altman’s and musk’s and zuck’s shit mouths, and anything you have read in any of the major media publications. As is tradition.