It’s crazy how most computer science degrees just focused on coding and now all of that is worthless. If they actually focused on learning how computers work instead of Java8 bullshit these people would probably have a job still.
Computer science was always a fake field made to placate losers who couldn’t pass circuit analysis 101. Thankfully computer & electrical engineers will still fair fine.
I don’t blame the students though, it’s more the colleges and universities for coming up with this crap.
Incredible. Jensen is completing the circle.
- Bankers don’t like GPUs as collateral because the depreciation is unpredictable
- It’s unpredictable because a new GPU can obsolete an old one
- Jensen knows his own roadmap
- so he’s offering depreciation insurance to the banks
- the depreciation insurance (up to 25%) helps the banks get marginal deals over the line
Speculation
- Nvidia will also advise the banks on “reference designs” for datacenters that will make them fungible
- Having them be fungible means that the debt can repackaged into Asset Backed Securities, Collateralized Loan Obligations and Collateralized Debt Obligation (ABS, CLOs and CDOs from 2008 haha)
- This allows tranching to get investment grade ratings on the debt so that it can be resold to pension funds and insurance firms
- It also allows the banks to trade idiosyncratic project specific credit risk for sector wide credit risk
So Jensen is trying to get his customers the same cost of financing as real estate rather than venture equity.
This is going to move the data center game out of the VCs and into the big leagues.
Newly passed MSME bill is unlocking ₹8L crore stuck in delayed payments, including many startups
If even 25% were timely rather 6 months of chase, would unlock 2L Cr of working capital, equal to 3x India's VC funding
Largest invisible tax on entrepreneurship being lifted
Everyone is shorting SaaS. Everyone is shorting IT consulting companies. Everyone’s seen what’s happened to Salesforce, Atlassian, & Infosys. The trade is crowded, the multiples have already compressed, and the alpha is gone.
But.. what if.. I told you there was still a bigger better play. One where a $3.7T economy is structurally short the labor arbitrage that frontier AI models are dismantling, and that the macro plumbing amplifies the shock instead of absorbing it…
I love my girlfriend.. but guys.. it’s time to short the Indian rupee.
India runs a structural current account deficit which means they spend more on goods/services from the rest of the world than they earn selling their own. They’ve run one every year since the economy liberalized in 1991. Running a current account deficit isn’t fatal, but it’s like a treadmill, you have to keep finding dollars to fund it, year after year, and if any one of the funding sources dries up, the currency adjusts. India has been running on this treadmill for thirty years. The reason India has been able to pull this off is because their deficit is funded by remittances from the Indian diaspora, FDI/portfolio flows, and critically for this trade services exports.
Services exports from India are now ~$370B annually and growing, and they have become the single largest plug for India’s external accounts. Goods exports are ~$440B but with imports of ~$680B, the goods trade deficit alone runs ~$240B. Services exports cover that gap and then some, which is why the INR has been remarkably stable around 83–88 per USD for years despite the goods deficit.
The composition of these services exports is why this is an AI trade. Roughly 55% of India’s services exports are IT service, software services, and back-office functions that the giant Indian consulting companies sell to US and European enterprises. That’s ~$200B+ annually. If you then add in Global Capability Centers (which are when international companies create an Indian office to offshore work. This is estimated to employ 1.9M people and is growing at 11% YoY), you’re looking at the entire IT services economy producing somewhere between $240–280B of annual export earnings, all of which is dollar-denominated, all of it billed by the hour.
This whole sector alone sustains India’s current account deficit and is in the direct line of fire to be automated by AI productivity gains. This is not a 5% problem. This is a 30–60% revenue compression over five years. Roughly half of the $280B base is routine software work like app development, maintenance, and testing that is already replaceable by GenAI coding tools. Another ~25% is call centers, claims processing, document review and that’s even more exposed because the work is more structured and the AI tools there are more mature, putting another ~$38B at risk of a 50% compression. The remaining ~25% is higher-value enterprise work that’s defensible for now but maybe ~$7B is at risk.
If you Stack them all you get ~$80B of annualized revenue compression by year five, or 29% of the base gone and that’s the floor for my estimate..
The bear case destroys $140–170B of service exports or 50–60% of the entire industry. The timeline matters here a lot as well. 2027–2029 the Fortune 500 companies will have a good idea of token usage and capacity rationalization leading to vendors shrinking and multi-year contracts repriced lower. Sadly, this exactly when India needs the funding most, because manufacturing PLI revenue won’t have ramped enough to compensate.
A 29% revenue compression means India’s IT services exports shrink from ~$280B to ~$200B annually with $80B of dollar earnings gone. The current account deficit today at -1% of GDP is ~$40B today. Meaning you don’t need my bear case to break the rupee just losing $80B of services revenue alone takes the deficit to roughly -3% of GDP, which is historically the level where INR has been forced to depreciate sharply.
I keep wondering if this dude will actually retire a hero or just roundtrip most of it, because when it comes to trading, there are very few people who can do what he just did, VERY impressive, but almost nobody escapes the ego trap that comes after.
The only escape is to retire from the exact sector where you won the most, and that is a brutal ask when you are young and you just won your first Super Bowl. Why would you leave the field right when the world is confirming you were right?
As always the hardest trade is exiting the identity.
Also, I see very few people talk about the non zero chance that this whole AI boom just ends up not being a good financial model, and then things crash.
We have seen this before across history, where the world gets excited about some new infrastructure like railways, the electric grid, or highways, builds into it very aggressively, and then the economics end up breaking.
I think there is a non zero chance data centers are the next version of that, the next global excitement where everyone rushes in, too much gets built, and that ends up bankrupting the economy.
And if that is the case, then Leopold probably only has 3 paths. He either does a complete 180 on his take to keep winning, retires, or doubles down and roundtrips all of it.
If the monopoly —> duopoly —> oligopoly —> commodity cycle can happen in the most advanced areas of technology, in part accelerated by AI, what is the true value of “long term” cash flows?
Long term may not mean what it used to…
$GOOG is so dumb. They paid SpaceX 4x the average rate for GPUs when they could’ve waited for $META to sell at liquidation prices
I may buy my own cluster when the neocloud bankruptcy auctions start
In the end, AI will produce more bloated products for us to troubleshoot and fix.
If you are a good software troubleshooter who understands the fundamentals. you will be on high demand. And Your skillsets are deadly.
the unemployed 23 year old gambler obsessed with markets and -$20,000 in the hole has better odds at making it than the employed 23 year old with no debt and a steady paycheck who thinks he got it all figured out