South America didn't have a mix of domestic and foreign investors deploying massive quantities of private money into capital assets in the 60s and 70s. They had governments borrowing to fund their citizens' consumption. Massive difference on multiple levels.
I'm not saying "this time will be different". I'm saying this is business as usual.
Zuckerberg lost $30bn or more trying to create a VR amusement park. Scale that up to $500bn and see how much waste and dead-weight losses are created.
"Up to $500bn" is business as usual for Silicon Valley post-2021.
2. There is no commitment to spend in a single year
3. There is no actual contractual commit here, this is a press release (i.e. Marketing)
4. There is not actually a $500B pile of gold being spent. This is more of a "this is how big we think this industry will be and how much we may spend to get exposure to that industry"
[1] https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile...
As Elon said, they don't the money.
Talk is cheap.
$500B is not.
The problem here being that it was money spent that was never earned back, and money that eventually had to be paid back, right?
This can also happen with private capital. 2008 was a bust caused by private banks, for example. AI hasn't proven to be profitable yet [1], and I'm not sure it'll makes a difference, for the success of projects like this, wether the money is coming from government or not.
In fact, if the 2008 bank bail-out, auto industry bail-out, the Silicon Valley bank prop-up, and other such actions by the US government are considered [2], if this turns out to be a bubble it will be taxpayers who end up fronting the bill.
[1] https://www.cbc.ca/news/business/ai-generative-business-mone...
[2] https://www.investopedia.com/articles/economics/08/governmen...
In part. It was money borrowed by the state. That means when it can't be paid back, it's automatically a systemic issue. And it was money borrowed to fund consumption. There was no good reason to ever expect it to be paid back because it wasn't funding productive activity.
> if this turns out to be a bubble it will be taxpayers who end up fronting the bill
Very possibly, particularly if part of the package are e.g. federally-subsidised loans. Before that, however, private parties will almost certainly lose tens if not hundreds of billions of dollars. That cushion, together with those parties being spread between domestic and foreign sources, is what makes this less risky to the United States than similar relative-magnitude projects in South America. (Plus the fact that this is a capital asset versus consumption.)
Haven’t all three examples you note (2008 crash, auto bailout, and SV prop up) resulted in a net return/gain for the taxpayer?
Since when does "private capital" speak in such honeyed tones to state powers?
https://www.cnn.com/2025/01/21/tech/openai-oracle-softbank-t...
Groupthink capital, directed by mostly 2 "thought leaders".
Economies don't like Groupthink Capital, regardless of it being private, public or a combination of the 2
Of course the US economy as a whole is huge so even billions can be absorbed, once you start talking about half a trillion though...
The demand for more AI compute is already here and is less risky of an investment.
"Centralized planning" was effective under Bell Labs
It's private money. CEOs will say whatever they need to say to achieve goals (here, favorable conditions for AI work), look at what the actual money flows say.
$500B is a lot - no matter how you slice it.
It's a lot easier to talk pie in the sky than to actually get $500B to spend.