> “But from a global allocation perspective, actually, it’s quite terrible to have everybody have a bunch of the world’s limited supply of DRAM sitting in their own computer and being used 5% of the time,”
It’s terrible from a global resource utilization perspective for people to fly around in private jets; but I can’t imagine Mr. Altman does otherwise.
So thank you for trying to protect us from the terrible scourge that is personally-owned DRAM, but I’ll take my chances.
Maybe Altman can resurrect the SETI@home project, though I would rather have my PC work on finding things among the stars instead of answering dumb AI queries.
Tech company execs are among the stupidest people on Earth. Another article went around recently where the former CEO of Yahoo couldn't think of three positive words to describe her own child[1]. It's baffling that these people who probably put their pants on the wrong way around 50% of the time get as much airtime as they do. Truly breathtakingly stupid people.
I feel like he could sell this as a future where people have super computers at home and when they’re not running they’re generating income, he doesn’t seemed constrained by the truth, just promise everyone that you’re going to replace their jobs and that their home computer will provide for them. This universal basic compute idea here seems consumed with is not appealing to anyone outside of tech.
So a gig economy but for compute? The problem with this, it will be a race to the bottom and will not pay enough to support the people. Not to mention the energy costs and wear and tear on the compute parts.
I listened to the full podcast, and this is taken completely out of context. His point was entirely logical: if there’s scarcity, it’s important to maximize utilization of the scarce asset, and you do that by putting it in a data centre where it’s serving people constantly.
His point is the highest and best use of the chips is to be full of model weights predicting the next word 24/7 instead of maybe being in a laptop that's off or underutilized sometimes.
It just so happens that the highest and best use is also what makes him the most money.
Nevermind the fact that there was no RAM shortage before he and his buddies decided to hyperscale their word guessers.
It'd be a bit like if an almond farmer came to a town with plenty of water, used up all the water growing almonds, and then tried to argue that growing food for everyone is a higher and better use than townfolk taking baths or watering their vegetable gardens.
It's not just ideologically bankrupt utilitarianism, it's ideologically bankrupt self serving utilitarianism.
According to your basic economic theory, none of us should have ever gotten any inefficient personal goods if they can be pooled into more efficient clusters. We wouldn't get personal computers, because obviously renting time from a large computer is much more efficient. We wouldn't get personal transportation, or non-communal housing, or inefficient small businesses. Are you sure you want to argue this point with all its logical consequences?
Let me know when you're ready to move onto advanced economic theory, including what happens to efficiency and value when a tiny number of entities concentrate an in-demand commodity in their hands and refuse to share.
It does happen to ignore individual property rights and restricting the free market, as well as sapping away consumer welfare from the older market equilibrium to be fed to the frontier lab beasts through increased prices and through utilization by legacy consumers through the LLM mark of the transaction, so there is that.
This forum has a hard time championing anything that isn’t “logical” as if ethos and pathos are inferior for some reason.
The man caused a worldwide RAM shortage. Who gives a flip about context?
I’ll take lower utilization for on prem thanks. Should I give my solar panels to my power company so they can overcharge me too in the name of “efficiency”?
>Should I give my solar panels to my power company so they can overcharge me too in the name of “efficiency”?
That doesn't make any sense because solar panels hooked up to the grid already benefits from high utilization. However if there is some national energy crisis, and there's a bunch of solar panels sitting idle, it might very well make sense to at least connect them to the grid so the energy isn't wasted.
If we want to talk efficiency, the market and investors seem awfully irrational to me. The whole scarcity and costs skyrocketing for other proven and more valuable use cases is because of inefficiency imo.
Scarcity is because of AI demand. Data centers are set up to meet that demand. It’s quite rich to say “evil companies are taking up all the ram” when they are serving billions of users and you want to do the exact same thing for yourself.
Is there any evidence that we're short on compute for the AI demand? My impression was that many data centers were being planned to serve hypothetical future demand (with the assumptions that it'll multiply many more times while also not becoming much more efficient) and to facilitate model training, neither of which has much to do with the argument you're making.
It’s not quite that simple. A few small companies are speculatively buying up all the production capacity going on into the future. At the same time they aren’t actually using it all; a bunch is juat sitting in warehouses not doing anybody any good. And there are real questions about whether this represents any sort of consumer demand, or if it’s primarily fueled by billionaires’ desire to lay off all their workers and replace them with chatbots.
So we’ve got rent seeking behavior squeezing everyday people on both sides, and if anyone says, “hey, I don’t like how this is going”, your answer is to say, “So then set up a factory?”
That’s so tone deaf it makes “let them eat cake” sound downright empathetic by comparison.
> At the same time they aren’t actually using it all; a bunch is juat sitting in warehouses not doing anybody any good
Is there any source for this beyond online clickbait? I find it hard to believe that RAM is piling up in warehouses while simultaneously every AI company and every consumer is desperate for it.
But he's creating the scarcity. If someone corners the market for something simple like cars and creates the economic conditions where consumer cars now cost 500k, does that person also get to go on a podcast and say "normal people shouldn't own cars anymore because they're a scarce resource?" as if it's just an objective reality that they played no part in creating? Because that's exactly what Altman is doing here.
"Scarcity is because of AI demand. Data centers are set up to meet that demand. It’s quite rich to say “evil companies are taking up all the ram” when they are serving billions of users, and you want to do the exact same thing for yourself."
Yes. Altman was just pointing out something about economies of scale.
You don't have to agree with his values (e.g. you could believe that individual ownership of computation is more important than overall efficiency) but the statement here is logical and more or less obvious.
An argument can be true in its parts and still mislead by presenting an oversimplified model. The quote below is pointing out that timesharing can obviously be more efficient in theory -- but ignoring the equally obvious objection that data center operators can't be trusted to allocate towards global efficiency in practice.
> A lot of people want local inference. And somehow, they assume that the problem is solved now. Because, we are going to have open-source models. We’re going to be able to run them on our computers. And so we’re not going to need all the stuff in the data center. But from a global allocation perspective, actually, it’s quite terrible to have everybody have a bunch of the world’s limited supply of DRAM sitting in their own computer. And, being used 5% of the time. You want maximum utilization in the cloud if you’re trying to deal with this very supply constrained environment. And somehow everyone’s... Many people’s intuition is the opposite of that.
Yes and as a side effect that guy gets unprecedented level of control over who gets to use that RAM for how much and what models will be used and so on.
Even things like decommissioning an old model can be extremely disruptive if all your workloads are tweaked for it. Like when Opus 4.8 dissappear from copilot in one company I work for. Suddenly our PR reviews started missing issues they always caught before.
So not very out of context at all? People see through it and are upset because the message effectively comes down to "We, the rich/corporations, deserve it more than you so we can rent it back to you".
I understand from other discussions that Sam has several houses in SF alone (plus other cities). He did not discuss maximizing the utilization of those?
I’d be more impressed with this scarcity argument if all the RAM companies are buying up were actually getting put online, instead of sitting around in warehouses as currently seems to be the case.
If you are the cause of artificial scarcity due to hoarding, you don’t also get to claim to be the solution to the problem. That’s just running a racket.
Maximizing worldwide memory utilization isn’t generally something that people (individual human people) care about. They usually care more about ownership.
Try swapping the subject for something less divisive, like cars. If some calamity hit where we can no longer makes sense to produce cars, it would certainly make more sense to reallocate everyone's private cars (with compensation) to taxi services so each car can be used most efficiently, rather than sitting on a parking lot 23 hours a day.
The problem is that the RAM shortage isn't caused by some "calamity", it's being financially manufactured by companies like OpenAI. The more accurate analogy is that OpenAI starts buying all the cars, jacking up the price on consumer vehicles, and then going on a show acting like he's not influencing market conditions at all.
>it's being financially manufactured by companies like OpenAI
How is this a relevant factor as long as the demand is genuine and not some attempt the corner the market? Otherwise blame all the people/companies buying the tokens.
While it is out of context, there is truth to his statement. There is something vile about people who become very rich very fast and they lose all sense of self-awareness and humanity. He wants all the resources for himself and his company. We're back to the "You'll own nothing and be happy" statement.
That would be true if the only use case of RAM was to hold context for LLMs or other cloud based processes. There are many other things RAM is useful for, not only running processes in the cloud.
Or should, for example, gamers all have to stream video games from Geforce Now instead of having their own PC?
How is that not taking things out of context (or rather, leaving them stripped of it)? Scarcity caused by what? For the benefit of who? Utilization for what purpose? "Serving people", heh.
I wouldn't want RAM or a GPU for a decent price to run local models, but to do normal computer stuff. That is what is scarce, RAM and GPU for people. Nobody would give a flying fuck if ONLY data centers had trouble acquiring RAM.
If someone fuckers buys up all the drinking water to make enemas 24/7, and people survive on the rain water they catch in bowls, and then those guys says actually, we should have those too, they're just greedy sociopaths, no need to complicate it.
Hospitals don’t use generators at 100% capacity so we should probably stop using them there /s
This is not logical because RAM a resource we want to use. There are many uses of RAM that don’t involve LLMs, and to assume that “max utilization” and “max utility” are the same thing is naive at best. The entire reason there’s scarcity is a problem of OpenAI’s own doing (along with the other labs).
Everyone is making fun of him for this clumsy take, but there is some truce to it. It's essentially how Cloud computing came about. We will have good LLM models that run on local hardware, and that is fine. However, it is foolish to think that it will ever make economic sense to run frontier models on hardware that sits idle 40% of the time.
Indeed. It is like cars vs public transport. Or private planes vs airlines. I am sure Vampire Sam would love to hand over his cars and jets because it is more efficient use of resources to take the bus.
I agree, if I am saying or doing cartoonishly evil things without any guardrails and with reckless abandon I too hate it when people call it out directly.
> from a global allocation perspective, actually, it's quite terrible to have everybody have a bunch of the world's limited supply of DRAM sitting in their own computer. And, being used 5% of the time. You want maximum utilization in the cloud...
I promise, if you ever have the misfortune of having an article like this written about you, your reaction will not be "yeah, that's basically what I said"
We will give business men a pass for being disingenous constantly, bowing down to fascists if it gets them a little more money, writing laws to enrich themselves more at the expense of the populace, bribing politicians so they can get a few more dollars.
But if a journalist gets a LITTLE snarky about any of that, oh goodness me how irresponsible!
Journalism has a stricter moral code. The whole point of journalism is to spread correct info. The point of business (most of the time) is to make money. These are very different. Also, since when did we give business men a pass?
> “But from a global allocation perspective, actually, it’s quite terrible to have everybody have a bunch of the world’s limited supply of DRAM sitting in their own computer and being used 5% of the time,” he added. “You want maximum utilization in the cloud if you’re trying to deal with this very supply-constrained environment. And somehow many people’s intuition is the opposite of that.”
Altman not acknowledging that companies like his are creating the conditions that cause the supply constraints is the Dr. Evil cherry on top.
Is it possible to reinvent memory with stuff everyone owns or a separate system? It's been a while since I did hardware and low level stuff, so approach this question from a more "inventive" and perhaps "dreamy" way. Or just educate me on physics again and tell me to stop being dumb.
I would be really interested to know how LLMs are being used globally, how much of it is scientific computing, coding tasks etc vs mundane questions people would have googled (and be satisfied with) 3ish years ago.
I'd argue its probably lower than anticipated unless you count Google's AI search mode in the LLM column instead of the search column. As a forced inject, it could go either way.
It frustrates me that these AI companies’ hardware grab is aggressively antagonistic to technology enthusiasts.
I find it reassuring that the leadership confirms that their intention is to make my hobbies and profession less financially viable, as I would otherwise worry I’m overreacting.
It’s terrible from a global resource utilization perspective for people to fly around in private jets; but I can’t imagine Mr. Altman does otherwise.
So thank you for trying to protect us from the terrible scourge that is personally-owned DRAM, but I’ll take my chances.
Who said the above? Was it:
1) the head of the public transportation department in 1930?
2) the CEO of IBM in the 1970's?
3) Sam Altman in 2026?
One with RTX5090, 64GB DDR5, Ryzen 9, 4TB goes for equivalent to $13500, with most of that being the GPU and RAM.
A year ago, the same PC would have cost a third or so.
In any case, with the combination of expensive GPU, RAM, who the hell can actually afford this stuff.
https://setiathome.berkeley.edu/
- https://github.com/exo-explore/exo - https://machinelearning.apple.com/research/distributed-gpu-c...
[1] https://www.nytimes.com/2026/10/25/style/marissa-mayer-ai-as...
Then don't buy the hardware in the first place and leave it up to the hyperscalers?
His point is the highest and best use of the chips is to be full of model weights predicting the next word 24/7 instead of maybe being in a laptop that's off or underutilized sometimes.
It just so happens that the highest and best use is also what makes him the most money.
Nevermind the fact that there was no RAM shortage before he and his buddies decided to hyperscale their word guessers.
It'd be a bit like if an almond farmer came to a town with plenty of water, used up all the water growing almonds, and then tried to argue that growing food for everyone is a higher and better use than townfolk taking baths or watering their vegetable gardens.
It's not just ideologically bankrupt utilitarianism, it's ideologically bankrupt self serving utilitarianism.
Something being self serving doesn't make it false, and you don't have to rabidly deny basic economic theory just because you hate the guy.
Let me know when you're ready to move onto advanced economic theory, including what happens to efficiency and value when a tiny number of entities concentrate an in-demand commodity in their hands and refuse to share.
(I say some of this tongue in cheek).
Does he actually call for expropriation without compensation? That part just feels like people massively extrapolating from what he said.
Why again are we spending all our money deploying AI when the AI-builders insist that this is going to result in biblical levels of unemployment?
The issue is that it can be applied to everything else, yet it isn't.
We can maximize utilization of wealth if we distribute it.
We can maximize utilization of land if we build more communal pools/third places instead of individuals each building their own.
We can maximize utilization of transport if we all stopped having our own cars and used public transport.
We can maximize utilization of everything if it was available to everyone at all times.
Humanity just hasn't been interested in maximizing utilization, and I certainly doubt that Altman is. That is the crux.
The man caused a worldwide RAM shortage. Who gives a flip about context?
I’ll take lower utilization for on prem thanks. Should I give my solar panels to my power company so they can overcharge me too in the name of “efficiency”?
That doesn't make any sense because solar panels hooked up to the grid already benefits from high utilization. However if there is some national energy crisis, and there's a bunch of solar panels sitting idle, it might very well make sense to at least connect them to the grid so the energy isn't wasted.
So we’ve got rent seeking behavior squeezing everyday people on both sides, and if anyone says, “hey, I don’t like how this is going”, your answer is to say, “So then set up a factory?”
That’s so tone deaf it makes “let them eat cake” sound downright empathetic by comparison.
Is there any source for this beyond online clickbait? I find it hard to believe that RAM is piling up in warehouses while simultaneously every AI company and every consumer is desperate for it.
All the RAM belongs to me is not too far off…
"Scarcity is because of AI demand. Data centers are set up to meet that demand. It’s quite rich to say “evil companies are taking up all the ram” when they are serving billions of users, and you want to do the exact same thing for yourself."
You don't have to agree with his values (e.g. you could believe that individual ownership of computation is more important than overall efficiency) but the statement here is logical and more or less obvious.
> A lot of people want local inference. And somehow, they assume that the problem is solved now. Because, we are going to have open-source models. We’re going to be able to run them on our computers. And so we’re not going to need all the stuff in the data center. But from a global allocation perspective, actually, it’s quite terrible to have everybody have a bunch of the world’s limited supply of DRAM sitting in their own computer. And, being used 5% of the time. You want maximum utilization in the cloud if you’re trying to deal with this very supply constrained environment. And somehow everyone’s... Many people’s intuition is the opposite of that.
Even things like decommissioning an old model can be extremely disruptive if all your workloads are tweaked for it. Like when Opus 4.8 dissappear from copilot in one company I work for. Suddenly our PR reviews started missing issues they always caught before.
If you are the cause of artificial scarcity due to hoarding, you don’t also get to claim to be the solution to the problem. That’s just running a racket.
How is this a relevant factor as long as the demand is genuine and not some attempt the corner the market? Otherwise blame all the people/companies buying the tokens.
Or should, for example, gamers all have to stream video games from Geforce Now instead of having their own PC?
How is that not taking things out of context (or rather, leaving them stripped of it)? Scarcity caused by what? For the benefit of who? Utilization for what purpose? "Serving people", heh.
I wouldn't want RAM or a GPU for a decent price to run local models, but to do normal computer stuff. That is what is scarce, RAM and GPU for people. Nobody would give a flying fuck if ONLY data centers had trouble acquiring RAM.
If someone fuckers buys up all the drinking water to make enemas 24/7, and people survive on the rain water they catch in bowls, and then those guys says actually, we should have those too, they're just greedy sociopaths, no need to complicate it.
This is not logical because RAM a resource we want to use. There are many uses of RAM that don’t involve LLMs, and to assume that “max utilization” and “max utility” are the same thing is naive at best. The entire reason there’s scarcity is a problem of OpenAI’s own doing (along with the other labs).
It is basically what he said.
But if a journalist gets a LITTLE snarky about any of that, oh goodness me how irresponsible!
Altman not acknowledging that companies like his are creating the conditions that cause the supply constraints is the Dr. Evil cherry on top.
2030 is soon btw ;)
I find it reassuring that the leadership confirms that their intention is to make my hobbies and profession less financially viable, as I would otherwise worry I’m overreacting.