Harry Potter III

the market can stay horny longer than you can remain salty

[Harry Potter III @netfuturevalue]

@Railway get your shit together!🥺

[Harry Potter III @netfuturevalue]

@NorwegianBYOB @nejatian You are very angry

[Harry Potter III @netfuturevalue]

Dj Khaled on his phone the entire game. What a waste of a court side ticket.

[Harry Potter III @netfuturevalue]

#djkhaled @djkhaled Dj Khaled sitting front row and on his phone the entire time!! Wtf

[Harry Potter III @netfuturevalue]

@TaraBull La Crack Connoisseur

[Harry Potter III @netfuturevalue]

@HarveenChadha It’s to disincentivize you using older models.

[Harry Potter III @netfuturevalue]

@MLCatttt @emollick Em dash (—)

[Harry Potter III @netfuturevalue]

@FreightAlley @grok which publicly companies are the biggest beneficiaries of the Supreme Court announcement on freight brokerages? Rank them by the biggest to lowest, considering their business/segment mix.

[Harry Potter III @netfuturevalue]

@scotus_wire @grok are there any publicly-traded, scaled freight brokerages that are going to be the beneficiaries of this announcement? $CHRW? $RXO? Help quantify the boost to their earnings. For this exercise, your role is of an experienced investment analyst.

[Harry Potter III @netfuturevalue]

@soobin @ReturnsJourney @grok answer this

[Harry Potter III @netfuturevalue]

@soobin @ReturnsJourney You lack the basic understanding of the three financial statements. Net margin belongs in income statement. Capex from AI statement belongs in cash flow statement.

[Harry Potter III @netfuturevalue]

The Montgomery Case That Threatens Brokers Is The One That Saves The Largest Broker

The freight industry is calling Montgomery v. Caribe Transport II an “extinction level event.” Thirty to fifty percent of U.S. freight brokers gone in a decade if the Supreme Court rules against preemption. Insurance demands hitting $5–10M. Broker fees jumping 15–25% overnight.

That’s the surface. Look at who actually loses.

The case asks whether brokers can be sued in state court for negligently selecting unsafe carriers. C.H. Robinson ($CHRW), the largest broker in the country, is the named respondent. Headlines frame this as Robinson fighting for survival.

Run the math on compliance. Vetting carriers at scale has a fixed cost floor. Data subscriptions for safety scores. Compliance staff. Legal review on every carrier onboarding. Insurance underwriters who actually want documentation.

For a broker doing $5M in annual revenue, layering on $200K of insurance and $150K of compliance staff drops margins from 15% to 7%. They can’t pass it through. Shippers route around them.

For C.H. Robinson, doing tens of billions in revenue across millions of loads a year, that same fixed cost is a rounding error. Their compliance stack already exists. Their data contracts already run at scale. Same regulation. Different outcome. Small brokers exit. The largest one books the loads they used to lose on price.

If preemption falls, Robinson loses the case and wins the market. If preemption holds, status quo, Robinson is still the largest broker. There is no version of this ruling where the industry’s #1 does not end up bigger.

The brokers panicking on X about extinction aren’t wrong about the body count. They’re misreading who benefits from it. The Court isn’t doing the killing. The unit economics of compliance are, applied to an industry that priced itself on the absence of compliance.

Watch what Robinson actually lobbies for between now and July. The intensity of their fight tells you whether they want preemption, or whether they’re letting a “loss” do the work for them.

[Harry Potter III @netfuturevalue]

@FreightAlley https://t.co/PWRXuYWp2s

[Harry Potter III @netfuturevalue]

The Montgomery Case That Threatens Brokers Is The One That Saves The Largest Broker

The freight industry is calling Montgomery v. Caribe Transport II an “extinction level event.” Thirty to fifty percent of U.S. freight brokers gone in a decade if the Supreme Court rules against preemption. Insurance demands hitting $5–10M. Broker fees jumping 15–25% overnight.

That’s the surface. Look at who actually loses.

The case asks whether brokers can be sued in state court for negligently selecting unsafe carriers. C.H. Robinson ($CHRW), the largest broker in the country, is the named respondent. Headlines frame this as Robinson fighting for survival.

Run the math on compliance.
Vetting carriers at scale has a fixed cost floor. Data subscriptions for safety scores. Compliance staff. Legal review on every carrier onboarding. Insurance underwriters who actually want documentation.

For a broker doing $5M in annual revenue, layering on $200K of insurance and $150K of compliance staff drops margins from 15% to 7%. They can’t pass it through. Shippers route around them.

For C.H. Robinson, doing tens of billions in revenue across millions of loads a year, that same fixed cost is a rounding error. Their compliance stack already exists. Their data contracts already run at scale.

Same regulation. Different outcome. Small brokers exit. The largest one books the loads they used to lose on price.

If preemption falls, Robinson loses the case and wins the market. If preemption holds, status quo, Robinson is still the largest broker.
There is no version of this ruling where the industry’s #1 does not end up bigger.

The brokers panicking on X about extinction aren’t wrong about the body count. They’re misreading who benefits from it. The Court isn’t doing the killing. The unit economics of compliance are, applied to an industry that priced itself on the absence of compliance.

Watch what Robinson actually lobbies for between now and July. The intensity of their fight tells you whether they want preemption, or whether they’re letting a “loss” do the work for them.

[Harry Potter III @netfuturevalue]

Chamath’s right, and the math is worse than the post suggests.

A 1GW data center needs 3-7 years to bring online. Permitting alone eats 18-24 months in most jurisdictions. Transformer lead times hit 120 weeks. Grid interconnect queues in PJM and ERCOT stretch past 2030.

So when you hear “powered land,” translate it: a permitted, energized site today is a 2029 capacity slot you can’t recreate by writing a bigger check.

Now the leverage math.

OpenAI signed for 4.5GW with Oracle in the Stargate deal. Anthropic locked 1GW+ with Amazon. Meta just committed $10B+ to a single Louisiana site. The hyperscalers aren’t buying compute anymore. They’re buying the right to exist at scale in 2028.

The people who figured this out two years ago own the scarce asset. Everyone else is negotiating against a clock.

Watch what happens next. The Amazon-Anthropic deal traded equity for capacity. That’s the template. Future rounds won’t price in dollars per token or revenue multiples. They’ll price in gigawatts delivered and the year they light up.

A model lead compounds for 6-12 months before someone catches up. A permitted gigawatt site compounds for a decade.
The foundational model labs spent five years thinking they were the scarce resource. They just found out who actually is.

[Harry Potter III @netfuturevalue]

@ewarren The em dash (—) is a giveaway for this AI slop.

[Harry Potter III @netfuturevalue]

Intel ($INTC) posted $13.6B in Q1 revenue against a $12.4B estimate. EPS came in at $0.29 against a forecast of $0.01. The stock moved 16%.

The headline is AI demand finally showing up in data center CPU orders. Q2 guide jumped to $14.4B. AMD ran 7-8% in sympathy and crossed $500B in market cap.

That’s the surface story. Here’s what actually happened.

The US government took an equity stake in Intel in 2025. That position is now up roughly 250%.

Read that again. The federal government isn’t subsidizing a strategic chipmaker. It’s a shareholder in one, and the shareholder just got paid.

The CHIPS Act was structured as grants and loans. Roughly $52B appropriated, disbursed as subsidies. Taxpayers got capacity, not equity.

When TSMC or Intel won, Washington got a press release.
The 2025 Intel stake flipped that math. The government now holds a position that marks to market every quarter. A 16% earnings pop doesn’t just help Intel’s cap table. It shows up on a federal balance sheet.

Three things change because of this.

One, the incentive to protect Intel commercially just got sharper. Export controls, procurement rules, tariff carve-outs: all of these now have a direct P&L line attached. Policy isn’t abstract strategic interest anymore. It’s a position.

Two, every other strategic sector gets a template. Shipbuilding, rare earths, pharma precursors, grid components. If Intel at 250% works, the next industrial policy package won’t be written in grants.

Three, the old fight about picking winners is over. The government already picked. The question now is whether it picks again, and on what terms.

Intel’s quarter was good. The 2025 stake was the actual trade.