Their IPOs are how they cash out and walk away. It’s just a matter of “does it crash before or after?”. Yes they’ll be anxious about getting there, but the fact the market is still behind them and they’re this close is very bad for the rest of us.
If they weren’t needlessly nervous (by that I mean paranoid), they would not be in that position.
Whatever their wealth, those people cannot feel safe.
Musk is never happy. He’s either chasing a ketamine high or getting rage baited on twitter. Real daddy issues with him. Zuck is just a hollowed out human.
I’ve heard (conspiracy) that the biggest offenders (meta/google) create shadow profiles for people who are not online. So its enough your picture or mention of you is online for them to be add it into the web of collectected data.
This has strayed quite a bit from the assertion that there can’t be enduring problems in the current crop of AI investment because it was at a high a while back.
But to continue, if you invested in the makeup of the DJIA in 1929 and left it in those companies, you would not be having a particularly good time.
If you kept refreshing the investment in the companies deemed to be the ‘good’ ones over time, then yes, your investment fared well, but that says nothing about the enduring value of the companies that made up the index in 1929. This is the current topic. It’s also wild to be this dismissive of the Great Depression as nothing to worry about because, eventually, we decided on a new set of lines that could go up.
So “line goes up” by definition is not really an endorsement that all companies lines go up, we shift the criteria over time to make sure that we measure whatever works in that moment.
if you invested in the makeup of the DJIA in 1929 and left it in those companies
Then you would no longer be investing in the DJIA as new companies were added.
But if you rebalanced periodically, you’d get the listed results.
So “line goes up” by definition is not really an endorsement that all companies lines go up
Diversification is one of the first lessons of practical investment. And blue chip indexes help simplify this process
If you switched from the DJIA to the S&P 500 or the NASDAQ, as these indexes emerged, you’d outperform the DJIA as it stood. I use the DJIA because it existed in 1929 while the others did not.
What’s stopping people from using open sourced trained models? Like, why are tokens going to be needed in the future? Isn’t this their whole business model?
Honestly, I have a little tin-foil-hat theory that the used their infine supply of venture to fuck the memory market to make sure people couldn’t run giant chinese models on solar and eat their lunch.
you mean like Sam Altman’s ‘dirty DRAM’ deal where he secretly negotiated with Samsung and Hynix to (together) buy 40% of the world’s memory production, announced both deals the same day (neither company knew about the other), and bought the RAM not as sticks or even chips but as finished wafers, useless until they are sliced, binned, and packaged into chip casings, a task OpenAI has neither the equipment nor the expertise to accomplish?
For enterprise, I think same reasons why companies using paid software instead of open software. Someone else to maintain and blamed if something doesn’t work.
training a top-level model takes harvesting petabytes / exabytes of revelant data and then months of continuous 100% power on a large server farm costing many millions of dollars.
It’s a one-and-done process, then the normal Q&A usage has vastly lower requirements.
Local models generally means you download the result of all that , after it’s done, (the PT in ChatGPT stands for pre-trained) and then generate answers locally, but the training wasn’t done locally. The generating answers step is called “inference”.
Hardware cost to do inference with a top-level model is within reach of most small businesses. Hardware cost to do training of a top-level model is orders of magnitude higher.
So, there are open-source models that are trained by universities, and then used by affluent but not billionaire people at home.
There are open-source models that can do inference (answer questions) on a typical 8GB gaming graphics card, but they aren’t in the same order of magnitude of quality as the famous top-level models.
Thank you! That’s the best answer I’ve seen on any platform!
What would you recommend as a local model? Want to dip my toes in but I’m woefully ignorant. And again, I’ve done my time in IT, so you don’t have to dumb it down, uh, too much.
The last time I ran one locally was a few months ago, so this might be somewhat out of date, but I’ll say Qwen.
Generally there are two numbers to look at first, the number of parameters (number of nodes + much larger number of connections between nodes) and the quantisation (number of bits that define each parameter)
So an 8 billion parameter model at 8-bit quantisation will require 8GB of working RAM, probably VRAM for fast results at home (at 8 bits to the Byte)
and also a 16 billion parameter model at 4 bits quantisation will also fill the same 8GB.
in my experience, high bit depth enables the thing to respond in more complex terms, and more parameters means more “knowledge” is there to be accessed. I have better results from phrasing questions that’ll have short answers and leaning more towards more parameters than bit depth.
The next number to consider is “context window”, which is basically how far back into your chat the thing can “remember” and have previously given information influence the next answer.
A year to two ago, when I was trying a lot of LLMs for coding tips, Google’s seemed the best for larger projects, simply because it had a much bigger context window than any other, and so could “consider” longer and more interconnected scripts as a whole, even if perhaps Claude was “smarter” with a small problem. I don’t know if this is still true in autumn 2026.
They’ve (commercial models) started keeping a per-account or per-chat log of important facts that exists alongside the context window, but doesn’t carry as much detail. Good for seeming to get to know you, not as good for writing new code that won’t conflict with a long-established bit of old code.
It’s not really practical to have million+ token context windows on home/gaming hardware.
(usually a token is a word, or a couple of tokens are a long or rare word, or a token is a punctuation mark)
LM Studio is good software with generally sensible defaults for downloading pre-trained models and running them on your own PC.
They’re the output of all that data centre processing. The model is fixed at that point and can’t “learn” anything more, so the pressure is always on to release a newer model that resulted from more data with better algorithms, but still small enough to be able to run on a consumer grade GPU
The hardware requirements are substantial. If companies or individuals want to run models locally, especially at the speed you’d expect from the big name models, then everyone needs a pretty serious GPU on their work station. That or the company needs to build and maintain a server capable of running whatever loads they expect to see.
None of this is impossible of course. But you pay for the tokens one way or another. Either you pay the upfront cost of the hardware and continuing cost of the electricity to run it, or you pay with the inconvenience of running a model very slowly on lower end hardware, or you pay a subscription for someone else to worry about all that.
This is why local models are still niche. Only hobbyists nerds are willing to go through the trouble of setting up and running a local model. Most companies still are not sure if AI is actually worth it for them. So it’s much easier to pay for tokens than the giant upfront cost required to build you own infrastructure.
If the numbers the AI companies are putting up as far as infrastructure cost vs revenue are anything to go by, it probably isn’t worth it to most companies.
They don’t need to, OpenAI and Anthropic are running at a huge loss to keep it cheaper to use them than to buy the comparable hardware required to replace them. The only way running open source models is cheaper is if you ignore hardware costs, or if you have a very specific and unusual usage pattern (i.e. constant 100% capacity, and not a clock cycle more or less), or if you go with another cloud provider who are also running at a loss.
If you want it well rounded I can see the argument for enterprise but for me personally I programmed one on my workstation and it’s trained on very specific things I think it barely uses up all 24gb vram I have available.
come on, I’ve trained toy size models too, you can’t seriously say they’re even vaguely approaching the capability of models that takes a whole massive server farm weeks or months to train.
If you do one simple task like I do in my job yes you can make a small one purpose built. Work in reading comprehension before commenting. If you re read which I know might be hard for you, you can see I said make it for a specific task. Example for you since you’re going to try and counter because you’re an idiot. I have to fill in the same multiple fields on an excel table for state documents. This is pulled from a sql database I literally have a model where it does that for me with natural language processing instead of me being a tool and using python exclusively when I already did that to make the easy simple tool. I suggest Amazon for reading comprehension before commenting books. If you message back I’ll send you some recommendations I tell to my students.
Right, and what proportion of enterprise use cases genuinely need an appreciable fraction of that power (because they are wide ranging and non-repeating)?
Anyone that says these things aren’t impressive is kidding themselves. But how do we get from impressive to productivity enhancing in excess of cost? I find smaller specialised models more plausible on that front, but jury is still out on the financials, we’ll find out I guess!
I’ll admit I’m experiencing a degree of pleasure in thinking about how nervous Anthropic and OpenAI must be with their IPOs supposedly coming up.
Their IPOs are how they cash out and walk away. It’s just a matter of “does it crash before or after?”. Yes they’ll be anxious about getting there, but the fact the market is still behind them and they’re this close is very bad for the rest of us.
Looks like the market is now trying to get out from under them.
They’re already billionaires many times over. Hard to feel nervous with that kind of wealth in your pocket
If they weren’t needlessly nervous (by that I mean paranoid), they would not be in that position. Whatever their wealth, those people cannot feel safe.
Idk about that. Musk and Zuck seem like very careless people.
Musk is never happy. He’s either chasing a ketamine high or getting rage baited on twitter. Real daddy issues with him. Zuck is just a hollowed out human.
Based on the current valuations of their companies. They will likely remain 0.1%ers, but it will still hurt.
Place your bets
Place your bets
They gotta pull Google and get into the advertising game. They are sitting on a gold mine of private data and I’m sure they know it.
The crazy thing is all that data is only good online.
I’ve heard (conspiracy) that the biggest offenders (meta/google) create shadow profiles for people who are not online. So its enough your picture or mention of you is online for them to be add it into the web of collectected data.
OpenAI already explicitly postponed their IPO somehow hoping for a better climate to IPO in…
A certainly… optimistic expectation.
You should not try to make sense of a decision they probably used chatGPT to make. It’s bad for your metal health
NASDAQ just crested a new high this month, before stumbling last week.
If the death of the tech industry is on the horizon, I can’t see it.
DJIA crested a new high September 3rd 1929 before stumbling 3% the next week.
If death of the market was on the horizon, they couldn’t see it.
The slide being attributable to some "I told you so"s coming to pass may bode ill for the bigger "told you so"s that come due in a few months.
But the market doesn’t act rationally, both going up and going down, so who knows.
And over a century later, what has become of it?
If someone bought $100 of stock in 1929, how much would it be worth today?
$100 of Krueger and Toll? $0
$100 of the entire DJIA on September 3, 1929 adjusted annually: $13,633
If you’d invested in January of 1929: $16,913
If you’d invested in January of 1920: $72,275
January of 1930: $31,708
It’s never gone to $0
This has strayed quite a bit from the assertion that there can’t be enduring problems in the current crop of AI investment because it was at a high a while back.
But to continue, if you invested in the makeup of the DJIA in 1929 and left it in those companies, you would not be having a particularly good time.
If you kept refreshing the investment in the companies deemed to be the ‘good’ ones over time, then yes, your investment fared well, but that says nothing about the enduring value of the companies that made up the index in 1929. This is the current topic. It’s also wild to be this dismissive of the Great Depression as nothing to worry about because, eventually, we decided on a new set of lines that could go up.
So “line goes up” by definition is not really an endorsement that all companies lines go up, we shift the criteria over time to make sure that we measure whatever works in that moment.
Then you would no longer be investing in the DJIA as new companies were added.
But if you rebalanced periodically, you’d get the listed results.
Diversification is one of the first lessons of practical investment. And blue chip indexes help simplify this process
If you switched from the DJIA to the S&P 500 or the NASDAQ, as these indexes emerged, you’d outperform the DJIA as it stood. I use the DJIA because it existed in 1929 while the others did not.
What’s stopping people from using open sourced trained models? Like, why are tokens going to be needed in the future? Isn’t this their whole business model?
You need 200GB of video memory :)
Honestly, I have a little tin-foil-hat theory that the used their infine supply of venture to fuck the memory market to make sure people couldn’t run giant chinese models on solar and eat their lunch.
you mean like Sam Altman’s ‘dirty DRAM’ deal where he secretly negotiated with Samsung and Hynix to (together) buy 40% of the world’s memory production, announced both deals the same day (neither company knew about the other), and bought the RAM not as sticks or even chips but as finished wafers, useless until they are sliced, binned, and packaged into chip casings, a task OpenAI has neither the equipment nor the expertise to accomplish?
Of course, why do you think they are running around with warnings about “rogue AI destroying humanity”?
They hope something of what they throw at the wall sticks and some legislations limit their competition.
By the way, this kind of throwing money at the wall unfortunately works (if you have enough money).
But also - with this kind of applications being useful, memory prices would jump regardless.
OK, I’m just guessing.
For enterprise, I think same reasons why companies using paid software instead of open software. Someone else to maintain and blamed if something doesn’t work.
training a top-level model takes harvesting petabytes / exabytes of revelant data and then months of continuous 100% power on a large server farm costing many millions of dollars.
That was my understanding. So how do local models work? (Been fucking with computers since the 80s, AI is all new to me.)
It’s a one-and-done process, then the normal Q&A usage has vastly lower requirements.
Local models generally means you download the result of all that , after it’s done, (the PT in ChatGPT stands for pre-trained) and then generate answers locally, but the training wasn’t done locally. The generating answers step is called “inference”.
Hardware cost to do inference with a top-level model is within reach of most small businesses. Hardware cost to do training of a top-level model is orders of magnitude higher.
So, there are open-source models that are trained by universities, and then used by affluent but not billionaire people at home.
There are open-source models that can do inference (answer questions) on a typical 8GB gaming graphics card, but they aren’t in the same order of magnitude of quality as the famous top-level models.
Thank you! That’s the best answer I’ve seen on any platform!
What would you recommend as a local model? Want to dip my toes in but I’m woefully ignorant. And again, I’ve done my time in IT, so you don’t have to dumb it down, uh, too much.
The last time I ran one locally was a few months ago, so this might be somewhat out of date, but I’ll say Qwen.
Generally there are two numbers to look at first, the number of parameters (number of nodes + much larger number of connections between nodes) and the quantisation (number of bits that define each parameter)
So an 8 billion parameter model at 8-bit quantisation will require 8GB of working RAM, probably VRAM for fast results at home (at 8 bits to the Byte)
and also a 16 billion parameter model at 4 bits quantisation will also fill the same 8GB.
in my experience, high bit depth enables the thing to respond in more complex terms, and more parameters means more “knowledge” is there to be accessed. I have better results from phrasing questions that’ll have short answers and leaning more towards more parameters than bit depth.
The next number to consider is “context window”, which is basically how far back into your chat the thing can “remember” and have previously given information influence the next answer.
A year to two ago, when I was trying a lot of LLMs for coding tips, Google’s seemed the best for larger projects, simply because it had a much bigger context window than any other, and so could “consider” longer and more interconnected scripts as a whole, even if perhaps Claude was “smarter” with a small problem. I don’t know if this is still true in autumn 2026.
They’ve (commercial models) started keeping a per-account or per-chat log of important facts that exists alongside the context window, but doesn’t carry as much detail. Good for seeming to get to know you, not as good for writing new code that won’t conflict with a long-established bit of old code.
It’s not really practical to have million+ token context windows on home/gaming hardware.
(usually a token is a word, or a couple of tokens are a long or rare word, or a token is a punctuation mark)
LM Studio is good software with generally sensible defaults for downloading pre-trained models and running them on your own PC.
They’re the output of all that data centre processing. The model is fixed at that point and can’t “learn” anything more, so the pressure is always on to release a newer model that resulted from more data with better algorithms, but still small enough to be able to run on a consumer grade GPU
The hardware requirements are substantial. If companies or individuals want to run models locally, especially at the speed you’d expect from the big name models, then everyone needs a pretty serious GPU on their work station. That or the company needs to build and maintain a server capable of running whatever loads they expect to see.
None of this is impossible of course. But you pay for the tokens one way or another. Either you pay the upfront cost of the hardware and continuing cost of the electricity to run it, or you pay with the inconvenience of running a model very slowly on lower end hardware, or you pay a subscription for someone else to worry about all that.
This is why local models are still niche. Only hobbyists nerds are willing to go through the trouble of setting up and running a local model. Most companies still are not sure if AI is actually worth it for them. So it’s much easier to pay for tokens than the giant upfront cost required to build you own infrastructure.
If the numbers the AI companies are putting up as far as infrastructure cost vs revenue are anything to go by, it probably isn’t worth it to most companies.
They thought they sold tokens, they actually sell marginally (6 months lead) improved models.
They don’t need to, OpenAI and Anthropic are running at a huge loss to keep it cheaper to use them than to buy the comparable hardware required to replace them. The only way running open source models is cheaper is if you ignore hardware costs, or if you have a very specific and unusual usage pattern (i.e. constant 100% capacity, and not a clock cycle more or less), or if you go with another cloud provider who are also running at a loss.
If you want it well rounded I can see the argument for enterprise but for me personally I programmed one on my workstation and it’s trained on very specific things I think it barely uses up all 24gb vram I have available.
come on, I’ve trained toy size models too, you can’t seriously say they’re even vaguely approaching the capability of models that takes a whole massive server farm weeks or months to train.
If you do one simple task like I do in my job yes you can make a small one purpose built. Work in reading comprehension before commenting. If you re read which I know might be hard for you, you can see I said make it for a specific task. Example for you since you’re going to try and counter because you’re an idiot. I have to fill in the same multiple fields on an excel table for state documents. This is pulled from a sql database I literally have a model where it does that for me with natural language processing instead of me being a tool and using python exclusively when I already did that to make the easy simple tool. I suggest Amazon for reading comprehension before commenting books. If you message back I’ll send you some recommendations I tell to my students.
Right, and what proportion of enterprise use cases genuinely need an appreciable fraction of that power (because they are wide ranging and non-repeating)?
Anyone that says these things aren’t impressive is kidding themselves. But how do we get from impressive to productivity enhancing in excess of cost? I find smaller specialised models more plausible on that front, but jury is still out on the financials, we’ll find out I guess!
IPOs upcoming? You’re being facetious, right?
Guess I need to move some money anyways.