I feel like one of the central arguments against communism is that the central government is too big to successfully plan an economy, especially when it doesn’t have the distributed demand signals of a free market.
At what point does a private enterprise become so big and well capitalized it suffers the same problem. There are some notions here that Jensen is a genius for being able to orchestrate the demand for its products.
But I suspect we’re going to discover that this is vulnerable to the same problem.
When the largest revenue companies do about 10x in sales what they're doing now. Current US GDP is about $33T. Total domestic revenue of the top 10 US companies is about currently about $3.4T. There are around 36mil US companies. The bottom 90% bring in about $3.5T in total revenues. This roughly fits the expected pareto (power-law) distribution . So, a corporate oligopoly doesn't seem likely anytime soon.
But communism doesn’t work in smaller countries either. I don’t think you have to be as big as the US government to have enough money to engage in deals that obscure demand signal.
I think this conflates the idea that - Sometimes conspiracies are boring. "I did a thing because of the boring, mundane case that it benefited me" is - while not sufficient for fraud - it is one of the building blocks.
And clearly one of the building blocks of cartels, monopolies, etc.
the point being "Boring and mundane" isn't a sufficient defense.
I think the biggest problem is the way Huang was creating demand by orchestrating the financing of the entire tech industry. The CFO in charge can handle the contract details, but do they have the strategic chops and relationships to keep doing that?
And it was all high risk. It may be only a once in a generation founder who wants to orchestrate that risk.
And again, no one was expecting NVIDIA to stay flat, massive growth was priced in.
Massive growth is no longer priced in. The P/E is 20, despite anticipation that this quarter’s results are 100% year over year.
I would say it’s under-priced, but I get really nervous when my thinking diverges this strongly from the market. Previous confidence in my thinking has not gone well.
I mean practically speaking if he gets hauled in front of congress we don’t know that money would keep flowing. Investor lawsuits, etc, like - it could have become financially unteneble as well.
Why does it matter that one of the key leaders of our aI strategy was having an affair with a Chinese spy while advising the president on our policy of selling chips to China?
1. Closed loop water systems use chemicals that can leech out of the system. There is 100% going to be more discoveries as time goes on with such a fast build out. Maybe not enough to make it not worthwhile? But not 100% safe either.
2. I’m anti-data center. I think at some point, you pull the levers you have? I think the tech industry is too consolidated and wields too much power, and the whole data center buildout has been done under a shroud of secrecy.
I think it’s rational to pull the levers you have, when the industry is predicting something close to the apocalypse.
Yes it’s mostly vibes, but until the industry stops acting like fucking nutters. I’m voting on vibes.
Can you privatize the losses if the banks are investing, providing debt to and repackaging that debt into broader instruments?
I worry that one of the reasons that Anthropic has 500 billion in commitments and everyone is running to the credit markets is to make it so that if they fail everyone suffers.
I mean sure, I agree with you in principal, but if not bailing out someone brings down the country or global economy? In a democracy that government is getting punished hard.
If the failure of a private company can bring down the country the government should either break it up or take it over immediately as a national security issue. Too big to fail means too big to exist.
The only reason this can happen is because of bailouts. When companies know they can be bailed out they will take far larger risks than they should. Instead governments should put policies and laws in place that prevent these types of problems from happening.
There's no obvious reason to reward people for failing. If it's necessary for the economy to not collapse, nationalize it, either wiping the shareholders out or paying the current (presumably low) price, and sell it later.
Well, in some ways that would mean the "running the government like a business" people would get their wish. They allowed the business to get so big, it failed, they are punished alongside the company for the failure.
If they can't privatize the loses, they should definitely not be allowed to privatize the gains. AI should be a public utility and be required to open source their models.
All the more reason to make them fail. FDIC should pay insurance and otherwise fail the banks. Capitalism cannot work if there isn't feedback that not watching for things like this will hurt you.
I actually liked the approach that happened with several large companies in 2008 - the stockholders get wiped out, the senior management is gone, but the depositors get protected.
Where did depositors almost get wiped out in 2008?
At NetBank? Only if they were over the FDIC limit. Stockholders got destroyed, though.
At Bear Stearns? No, depositors got protected in a convoluted deal. The shareholders got very close to nothing.
At Lehman Brothers? They did "break the buck" on money market funds for a bit, but weren't depositors made whole? And weren't stock owners wiped out in the bankruptcy?
Washington Mutual? Depositors were covered by the FDIC.
Where were depositors "almost wiped out"? Where specifically? Where were stockholders bailed out?
> Where did depositors almost get wiped out in 2008?
“The [Deposit Insurance Fund] fell to the lowest point in its history, a negative $20.9 billion on an accounting basis, by year-end 2009.
…
To address this issue, the FDIC adopted a novel approach that required the banking industry to prepay its quarterly risk-based assessments for the fourth quarter of 2009 and for the next three years” [1].
Where did “significant amounts of cash” this was funded out of, which “banks were holding at the time,” come from? From the Fed and the bail-outs.
Without the bail-outs, banks wouldn’t have been able to fund the DIF and the FDIC would have faced the threat of insolvency.
No. The FDIC can borrow from the Treasury. Not the Fed. The Fed backstops’ banks liquidity. The FDIC backstops banks’ solvency inasmuch as it affects insured deposits.
We needed new laws in 2008 because the Fed’s backstop alone was insufficient. These banks weren’t just illiquid, they were insolvent. TARP gave them liquidity for “troubled assets.”
Sure. We're still waiting to hear how that translated into depositors taking losses and shareholders getting bailed out, though.
I mean, look, you can point to some losses by depositors - those over the FDIC limit, and when two money market funds broke the buck. You can also point to some instances where the rescues helped stockholders, at least compared to what their losses would have been. But the general trend was that shareholders took a bath, and depositors did not.
Are you talking about SVB? Because FDIC paid out much more than just insurance, it made sure everyone was whole. I think that set a very dangerous precedent.
No, it is not. The way the SVB et al. collapse was handled was highly unusual. FDIC guaranteed deposits in excess of the insurance cap, bailing out (wealthy tech and VC) depositors to the tune of 30%ish of the fund set aside for such purposes. The Treasury stood up a new facility to prevent contagion to other banks.
You cannot do this if the big Wall Street firms collapse.
> You cannot do this if the big Wall Street firms collapse.
I'm not sure what you mean by this. Wall Street firms as such don't have deposits; Wachovia, Washington Mutual, and a number of smaller but still large banks were handled in similar ways to avoid any loss of even uninsured deposits.
Perhaps you're under the impression that the FDIC paid out of its fund for those deposits? (e: Originally I claimed this impression was wrong, but in fact I was wrong.)
That was my claim, but it does seem like I was wrong. I'm going to have to figure out where I got this bad information from.
I do think that "Wall Street firms" in even the most expansive sense isn't generally understood to include large commercial banks. Some companies like JPMorgan are big enough in investment banking to kinda land on both sides of the line, but I've never heard anyone refer to Bank of America or Wells Fargo as a Wall Street firm.
Colloquially, I do feel that "Wall Street" is strongly associated with the banking system that people touch regularly, and not boutique funds and offices that they've never heard of. Might be a jargon issue.
Privatizing the losses doesn't mean the rest of us are going to escape without a scratch. It just means that we let fail the things that are going to fail.
What makes this so politically unpalatable is that having constantly bailed out the last few cycles, everyone is now operating on the assumption that they will be bailed out, so many entities who in a more perfect world would know better have done a lot of stupid things, like pensions that have a lot of bonds that will go bad if the AI bubble collapses.
Frankly we're looking at massive problems and upheaval at this point no matter what we do. But the advantage of not socializing the losses is that the signals of what needs to be changed go directly into the economy. Socializing the losses means that the signals get muted, confused, and mangled and people in the future end up doing the wrong things as a result.
Unfortunately, at this point I think the choice is between people clearly and forthrightly losing massive portions of their retirement as visible losses, or people deceptively losing massive portions of their retirement as the bailout creates massive inflation and their retirement is inflated until it has a fairly similar amount of losses. Unfortunately, it is a feature for the rulers that the bailouts obscure the losses and dampen the correct apportionment of blame, even as it destroys the underlying economy. Giving people clear, correct targets for their rage, the same process of correct pain apportionment the economy needs to become healthy again, would point people at the "wrong" clear, correct targets, and the clear, correct targets are the ones deciding whether or not a bailout occurs. Not just the literal lawmakers nominally making the call, but the entire system that makes the decision.
Plan on them trying bailouts. Every inch of us should resist.
We've been living in an economy distorted by the forces that lead us to this point. We've been locked into the decisions necessary to keep things going as they have been, whether or not they make sense.
The correction that you're talking about frees us to actually make choices. Some of them will be tough choices, but we finally get to make them, with all of our resources and work ethic intact, which don't just disappear because a stock index craters.
If we want trains and local farms instead of data centers and new iPhones, we can actually make that choice, instead of being forced into the latter because every investment fund is pumping them to the point of being a foregone conclusion.
To borrow an aphorism: the bust is a feature, not a bug.
I think the precedent folks were pointing to was the great depression--no strong signal of assurance to temper public concern means things like bank runs and sudden collapses of institutions that would have otherwise made it because of group fear. Not sure how accurate that comparison is though.
We do definitively know, with climate change level certainty (read: there is still the odd ball expert that dissents), that it absolutely needed to happen.
What is much less known, because it kinda kills the talking point, is that the "too big to fail" bailouts were fully repaid with interest, to the tune of an ~$17B profit.
The losses from the program were from the auto industry ($12B) and the struggling homeowners fund ($31B). Which is a bit ironic, because those two are probably the ones that would have the most public support.
Doomsday scenario is a self-reinforcing feedback loop of:
1) economy tanks, people earn less disposable income or become unemployed
2) consumers spending less tanks the economy even more
We have historical precedent for this in the great depression, and the government going hands-off back then made the problem almost certainly much worse.
You could even make a pretty credible argument that the Nazis only got into power thanks to this, so there's your doomsday scenario.
I don't think this characterizes the situation accurately. The economy doesn't cease to exist, it moves out of the existing financial system. Able-bodied workers, supplies, equipment, knowledge all still exist, they're just awaiting orders. You have your choice of how to spur productivity with the extant resources, including direct investment and propaganda. Many people credit WWII for the recovery, but it was actually just those two things, with the war as the excuse. But you can do it without a war, as with the Japanese economic miracle: some amount of austerity, everyone works towards a common goal, the government backstops the effort. Bonus: no golden parahute overhead.
I think we screwed up after 2008 and I would like to see us get it right this time. The government absolutely should be hands-on, but the focus should be on raw productivity and public well-being, not making certain that certain firms survive the bloodbath.
You know what's a great economic stimulus? You now own the house you owe a mortgage on, clear and free. "But that would bankrupt the mortgage-holding bank!" Okay.
I don't disagree with most of what you say, but one crucial point you miss in my view is that letting public trust in banks & "the economy" collapse is really "expensive" (in hard productivity, not just on paper).
This is what was allowed during the great depression and what we desperately tried to avoid in later crises.
Western democracies are not set up at all to let the government direct and compensate millions of workers on the turn of a dime; micromanagement by government in such a situation can be expected to be extremely wasteful and ineffective, and this only got worse, because we have less generic/unqualified use for human labor than in the past (construction, railroads, motorways, heavy industry), and specialists are even harder to manage.
If everyone stops even just believing in the ability of local supermarkets to provide toilet paper next week or, on the other end of the spectrum, starts distrusting the banking system/dollar entirely, then the consequences can be extensive, expensive and slow to recover.
It might even be worth it to feed taxpayer money to rich greedy capitalists because the public comes out on top still by avoiding that trust collapse.
That's fair, but I think you're again conflating trust in the overarching institution with trust in its current constituents.
This is important because it exposes the needle you have to thread: messaging that makes clear that you're intentionally throwing out the old banks and companies (or their leadership) in order to preserve their prosocial functionality without their antisocial dysfunction. People who are happy to see Goldman Sach burn still want a way to safeguard their money; people who want Mark Zuckerberg to pay the piper don't necessarily want to lose their Instagram galleries or Facebook connections.
The people who run these companies want us to believe that a collapse of the status quo is a total collapse, but it's not. And a government that facilitates transition of ownership and leadership to more capable hands doesn't have to follow the oligarch playbook; different ownership models and transition processes exist.
I'll give an example. Consider people living in a corporation-owned apartment complex. The company collapses - most such corporate landlords collapse - and maybe the bank holding the loan does too. There is no world in which the complex sells at its previous value, and there is probably a temptation to facilitate some transaction, any transaction, that would bring things back to order by designating a concrete owner. Of course, there are people living in the complex, and God knows what happens to them when God knows whoever shows up claiming rights to the units.
Or.
The government steps in. They claim eminent domain and set aside reasonable funds to pay whoever turns out to be the final previous owner. A co-op is formed and ownership is transferred to the current residents as a subsidy; they now own the units they previously rented. Their lives are not disrupted by forced eviction; the municipality is not disrupted by an explosion of homelessness. The main losers are the old owners and banks, and they're bankrupt and dissolved anyway. Life goes on.
I'm not saying that it's an easy transition. Russia showed us how to screw it up. But the successful scenario is a far better outcome for far more people than throwing our hands up, exclaiming that we don't know how to do anything without letting vulture capitalists have their way, and inviting them in to freely carry out the exact scenario (which continues to distort markets and makes the next reckoning all the worse) which is the failure mode of the above proposal anyway.
Some countries like Iceland let their banks fail (in no small part because their total debt was something like 10x their gdp), and they even successfully prosecuted a number of executives. They got moderate slaps on the wrist but did serve some token time in prison as opposed to having to pay a ground-breaking fine of 5 minutes of revenue as is more typical. It wrecked their economy in the short-term but they were back to growth by 2011, and had a larger than ever economy by 2015.
But I think more than the one-off situation is also the precedent it sets. When you can bank on the government bailing you out (pun intended I guess?), it creates a perverse incentive of just trying to get as large as you possibly can, all foundation or principles absolutely be damned - because once you get big enough you win, even if you're the biggest loser ever. And I do think that's what we're probably seeing with LLM companies.
Not that LLM's a loser - it's a revolutionary tech and will only become more so. But multi trillion dollar valuations for a product that has viable open competitors already available and whose operating requirements continue to plummet and will almost certainly continue to plummet? This is extremely amusing, but absolutely batshit insane. It's like paying millions for a pretty flower because it's novel in spite of the fact that you know, and everyone else knows, that once people start planting them there will be endless amounts available for negligible cost.
Anthropic is not going to fail. Claude does amazing things and all the companies built on top of it are more likely to fail because at the end of the day, we can remake their value proposition by using Claude and a few command line tools…
If Anthropic has made promises that its wallet cannot keep (as it looks like), it will fail. Let me remind you that Anthropic isn't profitable yet, and there's no clear path that it will be.
Okay fair enough. I don’t track the finances just what the tools are doing. And on that all these AI-Ifying documents and stuff doesn’t really have a moat, but Anthropic has a great product.
Either way the people should never bail out corporations.
To be fair: You'd need to make at least several billions of revenue by selling pencil sharpeners while mostly spending the $100b on R&D (not the production of individual sharpeners)!
The market for AI is very clearly there; people are willing to pay more than enough so that providing LLM infrastructure (inference) is profitable.
Just the SOTA model R&D is pretty unsustainably expensive right now, but this only means that if the status quo "bursts" (because people realize that pushing the frontier is not paying off), model improvement will (probably) be slower afterward.
Note: this could also happen because expensive SOTA progress fails to keep ahead of "garage-trained" models, in such a case SOTA-AI lab investments collapsing would not really slow AI progress anyway.
Because (some) investors with money still believe that this overspending (tbd.) on R&D is gonna pay off in the form of market share (or the generosity of our future AI overlords ;), and (some) dumb money is just following it.
An investment that needs to return 7x, alone, on hundreds of billions to trillions of investments... While not destroying the entire economy, or replacing it?
If you can capture a significant fraction of the gains made by automating lots of white-collar work, then those billions/trillions look more reasonable.
"Correct" investment volume towards early industrialization might have looked just as unreasonable without the benefit of hindsight.
great products only take them so far, the question is are they great enough so people would be willing to pay the prices they'd need to ask if they would want to at least match their spending with their income
Yeah Pro subscription. I realize building something targeting API is probably astronomically more expensive.
And I realize also that they are reducing churn by keeping the value there for subscribers. Not sure what that’s costing them, but it’s definitely holding me as a customer. The product works so well for me that I honestly don’t need to even glance at the competition.
And honestly, this has been the case since day one. I’ve always used Claude. ChatGPT a bit in the beginning, but it’s emoji fetish and “bro” personality gave me the ick so I dropped them.
> I don’t track the finances just what the tools are doing.
Then you have absolutely no idea what is going on. It has exactly zero to do with what AI can do, and everything to do with how the companies are being run. The technology will not die, but the AI bubble will pop, regardless of what Claude can do.
Jet airliners do amazing things, moving people across continents. But they are seldom profitable. Airlines, in general, struggle to make money and have to try every trick in the book to sometimes squeeze a profit.
What is this “everywhere else” you speak of. Tik tok is owned by Larry Ellison. Meta we’re talking about. Reddit? Is that smallest of platforms “everywhere else”?
Do you really believe this? Everything I hear from self described culture warriors like Musk was the left in the 80’s and 90’s was anti censorship as opposed to the moral majority on the right? You would find yourself in narrow company to say the left is fundamentally pro censorship, particularly in America.
Actually you'd find that almost everyone on the right believes the left is fundamentally pro censorship, along with other "facts", like the left is a hive mind of sheep, or the left is controlled by George Soros.
Which far right are you talking about? The lefts version of the far right where everyone to the right of Bernie sanders is Hitler? Or like Joel Webbon / Doug Wilson women shouldn’t vote far right? Who is this anti-free speech moderate right you describe?
Elon describes himself as moderate so is he one of these folks that doesn’t believe in free speech? He would say he stayed in the center and democrats moved left. What speech is this moderate right censoring?
Why yes. And as I predicted, I was censored. Meanwhile every leftist message left alone. It's a artificial echo chamber - created by people who ultimately hate freedom and free speech.
Our family escaped communism the hard way - it's grim how history repeats the worst.
And I sort of wonder what point you’re trying to make, because you could take two lessons away -
One, Trump may not have been reelected if meta hadn’t banned him. So maybe it was another case of meta trying to help republicans.
Or, if Trump hadn’t been reelected, maybe the republicans wouldn’t be facing one of the worst midterm routes in history.
But honestly I’m not actually sure the point you’re trying to make. Meta hired a Republican strategist into their c suite since then. Sooo much has happened, there’s a lot of historical context to try to put it in.
I feel like one of the central arguments against communism is that the central government is too big to successfully plan an economy, especially when it doesn’t have the distributed demand signals of a free market.
At what point does a private enterprise become so big and well capitalized it suffers the same problem. There are some notions here that Jensen is a genius for being able to orchestrate the demand for its products.
But I suspect we’re going to discover that this is vulnerable to the same problem.
reply