🛰️ Data Centres... In Space?

Every manager faces the same situations before lunch: a message to land, a meeting to run, a hiring call, a report due. The AI Report built 37 Claude prompts for exactly those moments, organised by the situations every manager faces. 

Copy the prompt, fill the brackets, run it in Claude, and get back 10+ hours a week. Oh, and it's free. 

All you have to do is subscribe to The AI Report, a 5-minute daily AI brief read by 400,000+ business leaders at IBM, AWS and Microsoft, and the full prompt pack lands in your welcome email. The newsletter and the prompts, both free. Subscribe and grab both

Subscribe and get 37 prompts free

SpaceX stock just sank below its own IPO price.

Shares dipped 1.2% to $134.50 on Thursday, sliding under the $135 debut price from last month.

Four days after listing, the stock hit $225.64. It's now down roughly 40% from that peak. Not even a fast-track invite to the Nasdaq-100 could stop the slide.

So what's spooking everyone? Two things: maths and chips.

Musk says SpaceX could be pulling in $1 trillion in annual revenue by 2030.

Current reality: $18.67B in revenue in 2025, with a $4.94B net loss.

So the pitch is roughly 53x revenue growth in four years, from a company currently losing $5B a year. Investors are doing the sums and backing towards the exit… quickly.

Enter Terafab, the mega semiconductor project SpaceX is co-funding with Tesla. Oppenheimer just called it "critical" to SpaceX's entire valuation.

The plan:

Oppenheimer's Timothy Horan is bullish, with an Outperform rating and a $250 price target. His logic: if Terafab works, SpaceX becomes the only vertically integrated AI company on Earth. And off it.

But even he admits the JV terms aren't finalised. And Musk's goal of 1 terawatt of chips per year (roughly 50x today's entire global advanced-chip production) gets filed as "directional" rather than a hard target.

Translation: nobody, possibly including Musk, knows the real number.

SpaceX is trading like a story stock, and right now the story has plot holes.

The bull case rests on a chip project with no finalised terms, a partner with zero fab experience, and brutal execution risk. Even TSMC $TSM ( ▼ 0.22% ), the best chipmaker on the planet, hit delays building its Arizona fab.

The near-term catalyst? Starship's Flight 13, which could launch as early as Thursday. A clean flight feeds the narrative. A fireball feeds the bears.

Oh, and some analysts think Terafab is step one towards a full Tesla-SpaceX merger. File that under "things that would break financial Twitter."

Wispr Flow turns your voice into clean, professional text inside any app. Emails, Slack, client updates — speak once, send without editing. 4x faster than typing.

Try Wispr Flow free

New York just became the first state in America to ban new data centers.

One year. Full stop. Anything over 50 megawatts is frozen.

And Trump is absolutely furious about it 👇

On Tuesday, Governor Kathy Hochul slapped a one-year moratorium on large new data centers. First state in the US to do it.

The reasoning: data centers are gobbling up power and straining water supplies, and New Yorkers are footing the bill. The state already has the 8th most expensive residential electricity in the country.

During the freeze, no new discretionary permits get issued. Meanwhile, the state will write up environmental standards for how these facilities get built and run.

Once the standards are done, the ban lifts.

Oh, and one more number: over 12 gigawatts of massive energy loads (mostly data centers) were queued up to connect to New York's grid as of May. That's a lot of demand suddenly left standing at the altar.

Trump went off on Truth Social on Wednesday, calling it a "terrible decision."

His argument: data centers are "Money Machines" that bring taxes and jobs, and New York just handed all of it to Alabama, Florida, Texas, and Arizona.

He also claimed the facilities "must pay for their own Water and Power," with leftovers going back to local communities. And he warned the freeze risks losing AI and data centers "to China."

Whatever you think of the caps lock, he's pointing at something real: the states that welcome data centers are hoovering up billions in AI infrastructure spend. New York just took itself off the menu for a year.

The AI buildout is the single biggest capex story in markets right now.

Google $GOOG ( ▲ 3.61% ), Meta $META ( ▲ 3.07% ), Microsoft $MSFT ( ▲ 2.78% ), and Amazon $AMZN ( ▲ 3.02% ) are pouring hundreds of billions into cloud and AI infrastructure. SoftBank and OpenAI's Stargate Project alone is a $500 billion plan to blanket the US in next-gen data center campuses.

All that money needs somewhere to land. New York just made sure it lands somewhere else.

The bigger signal: the backlash is spreading. Dozens of state legislatures have introduced bills to rein in data centers' impact on power bills and the environment. New York is just the first to pull the trigger. Maine's governor vetoed a similar freeze in April.

If more states follow, the AI buildout gets slower and more expensive. That touches everyone from Nvidia $NVDA ( ▲ 0.33% ) and AMD $AMD ( ▼ 3.47% ) (the GPU kings powering all this compute) to the hyperscalers writing the cheques.

For now, it's one state. But every utility bill that spikes near a data center creates another angry voter, and politicians can count.

The AI trade now has a new risk column: local politics.

The US is thinking about invading an island to win a war it's already been fighting for five months.

Not just any island. Kharg Island, the hub that handles most of Iran's oil exports.

The Wall Street Journal reports Trump is leaning toward expanding military operations against Iran after a Situation Room huddle with VP Vance, Defense Secretary Hegseth, Secretary of State Rubio and Joint Chiefs Chairman Gen. Dan Caine.

Two big options on the table 👇

Deploy ground forces to capture Iranian islands near the Strait of Hormuz. Kharg is the headline target, but the Journal says Trump's team is also discussing occupying other islands along the Strait to protect shipping and dismantle militarized territory.

One small problem: troops sitting on captured islands make excellent targets for Iranian missiles and drones.

Some US officials say Trump doesn't actually want to commit ground forces. He's already walked back his loudest threats before, including taking Kharg and going after Iran's oil.

The other idea is striking Pickaxe Mountain, an Iranian facility buried up to 475 feet beneath solid granite. That's deeper than the Statue of Liberty is tall, hiding under a mountain.

The US hasn't touched it yet. Trump, meanwhile, told radio host Hugh Hewitt: "We're going to take out Pickaxe Mountain."

He also says he prefers diplomacy and hasn't finalised anything. So take the mic-drop quote with a barrel of salt.

Why escalate now? Iran still won't hand over its nuclear stockpile despite weeks of strikes. Bombing hasn't worked, so the White House is shopping for new pressure points.

And the market already voted on what this means. When Trump first threatened Kharg Island, US crude popped above $90.

What it means for you: Kharg is Iran's main oil-export valve. Any credible move on it, or the Strait of Hormuz around it, puts a fear premium straight into crude prices. That flows into energy stocks, airline costs, inflation prints, and eventually the Fed's rate maths.

Escalation headlines are now an oil catalyst on their own. The threat moves the price before a single boot touches sand.

Watch two things: whether "Pickaxe Mountain" goes from radio soundbite to Pentagon operation, and whether tankers start rerouting again. That's your early warning system.