Comment by vintermann
6 hours ago
Too big to fail, nation state edition?
But yeah. I looked at how my pension funds are invested, and it's like 70% in the US. Even more than the sovereign wealth fund, which is apparently around 50% (if a quick AI search can be trusted).
Thing is, I'm not sure I will like the world where OpenAI and co. are as wildly successful as their valuation suggests either. So maybe I should invest in them, so that if it comes to pass, I at least have money?
> if it comes to pass, I at least have money?
I bet 20 GBP that Brexit would happen for that reason. Ended up with around 100 GBP to drown my sorrows :(
> Thing is, I'm not sure I will like the world where OpenAI and co. are as wildly successful as their valuation suggests either. So maybe I should invest in them, so that if it comes to pass, I at least have money?
I think their current valuation isn't suggestive of a world that's going to change very much.
If investors were taking seriously the idea that this could be "it" with AI enabling even full automation — not superhuman AI, not even a fast learner, just AI that can fully automate everything we've currently got and not be limited to the subset of desk jobs that LLMs can do OK — it would allow the economy to double in size in whatever wall-clock time period it takes for the AI to gather enough training data by simply observing human workers doing the things the AI has not yet learned to do.
If self-driving cars are a good example in this regard, that observation time may quite large:
Current AI has not yet mastered full self-driving of cars in general conditions, despite all the cameras on cars gathering data about how all the other (human-driven) cars around them behave in real conditions. To take a somewhat arbitrary cut-off points for the sake of illustration, going from the 2007 DARPA Grand Challenge to today in self-driving cars, would suggest a growth of 3.9%/year.
Global GDP growth of 3.9%/year would be worth a much higher valuation than the AI companies are getting, and yet still be slow change.
>Current AI has not yet mastered full self-driving of cars in general conditions.
I am inclined to agree but Waymo seems to provide a counterpoint? Is this because they have a good system for keeping human managers in the loop?
Waymo is the best, but IMO even they are not yet a counterpoint, because they're geo-fenced.
If I had to put money on who gets fully-general first, I'd put it on Waymo; but I have no idea what anyone in China is doing so there's plenty of room for surprises.
1 reply →
> f investors were taking seriously the idea that this could be "it" with AI enabling even full automation — not superhuman AI, not even a fast learner, just AI that can fully automate everything we've currently got and not be limited to the subset of desk jobs that LLMs can do OK — it would allow the economy to double in size in whatever wall-clock time period it takes for the AI to gather enough training data by simply observing human workers doing the things the AI has not yet learned to do.
This assumes that the economy and the governments will happily suffer the electricity and financial demands that Sam and co would give and the world would accept. Even if Sam's words were true, it is still a leap of faith because there is no data backing those words.
My take in late 2023 was that GPT-5 would either further push transformers into the "there is still a real chance transformers might be IT" or will become the unquestionable sign that transformers have peaked as a general technology. My take in 2024 was that the AI bubble was going to be noticed. My take in 2025 is that AI companies (other than the big ones like OpenAI) are going to struggle getting funding.
> This assumes that the economy and the governments will happily suffer the electricity and financial demands that Sam and co would give and the world would accept. Even if Sam's words were true, it is still a leap of faith because there is no data backing those words.
Most economic growth since electricity became economically relevant assumes that, the only thing that changes are the names of who (supplies and) demands the energy.
> Even if Sam's words were true, it is still a leap of faith because there is no data backing those words.
Ignore Sam's, assume mine in the hypotheticals of my comment: even something slower than Musk's ongoing failure to make FSD is still a massive increase over the status quo that justifies the price, yet it can do so in a way that's still not going to feel like a radical shift to live through.
> My take in 2025 is that AI companies (other than the big ones like OpenAI) are going to struggle getting funding.
Mine is that there's definitely a bubble overall, but not necessarily in any single player: who wins the pop and who disappears is unclear, but whoever does survive can easily reduce spending when there's no competition and they're consequently not having to put 1/3rd of their compute budget into the next marginal improvement needed to beat whoever else just beat them.
> just AI that can fully automate everything we've currently got and not be limited to the subset of desk jobs that LLMs can do OK — it would allow the economy to double in size in whatever wall-clock time period it takes for the AI to gather enough training data by simply observing human workers doing the things the AI has not yet learned to do.
That would nuke the economy, as it currently exists. What's any modern American business following McKinsey "best practices" going to do to the people whose jobs it can automate? It'll fire them all, ASAP (except maybe one or two left to do supervision and monitoring, if it's not totally brain dead). Then the entire mass-consumer-driven economy would collapse and wither, due to mass unemployment.
But the stock would go up this quarter, so it's a win that must be pursued.
That would happen even with UBI, because it's not like they'd replace your salary with free money. You'd have to shrink your lifestyle to live like the bottom quintile. You'd get just enough so you'd not be desperate enough to topple the system that no longer values you, and not a penny more.
And I think none of that's going to stop anyone. If AI gets to the point you described, our current economy will be destroyed, most people will be left behind, and the economy will reorient into a weird thing chiefly concerned with satisfying the whims of the riches billionaires.
> That would nuke the economy, as it currently exists. What's any modern American business following McKinsey "best practices" going to do to the people whose jobs it can automate? It'll fire them all, ASAP (except maybe one or two left to do supervision and monitoring, if it's not totally brain dead). Then the entire mass-consumer-driven economy would collapse and wither, due to mass unemployment.
I also said "not even a fast learner".
If everything is just automated, all immediately at once and the AI is forever able to just do all possible labour, rather than as I suggested a case where it takes the AI as long to gather sufficient training data from watching humans as self-driving cars are taking to do that, that nukes the (human relevance to the) economy.
But that's not the scenario I gave there.
This scenario I gave is one in which humans do still have a comparative advantage: we learn faster, so we can get working on new jobs and make a decent amount in each of them before any given AI has time to master the skills needed for that job's tasks by watching us.
(That said, and I didn't think of this with the previous comment, there's also a whole bunch of questions that start with "yes, but how much does it cost?" and then dive into a huge level of detail).
> You'd have to shrink your lifestyle to live like the bottom quintile.
Way ahead of you :P
Actually, while bottom quintile would be an increase for me (though an unnecessary one as I'm fine because I don't care about the materialist lifestyle), keeping even that would rely on my tenant still being able to pay me rent, so in any realistic scenario I'd still have problems.
> and the economy will reorient into a weird thing chiefly concerned with satisfying the whims of the riches billionaires.
It will definitely be weird, but I suspect most or all the current big winners are likely to be very unsatisfied. Outside Context Problem.
1 reply →
You can't invest directly in OpenAI because it's privately held. Many investors have been putting their money into OpenAI's suppliers such as Nvidia.
Clearly you just need to invest in any company open ai might express interest in as they all go up 20% on public announcements
I wouldn’t worry about being 70% of the US, especially since you have that 30% not in the US.
Many corporations are listed in the US even if they do a lot/most/all of their business overseas. Heck, there are Chinese companies that sell nothing in the USA listed on the US stock exchanges like XPENG.