Bluechip

$250 billion of orders for $75 billion of stock. One fixed price, $135 a share, take it or leave it. 555 million shares. Roughly 95 times sales. Five 2X leveraged $SPCX ETFs launch the first morning it trades, and the passive index machine is forced to buy $15 to $30 billion more through a float under 5 percent.
This is not price discovery. It is queue discovery.
Nobody is being asked what SpaceX is worth. They are being asked to line up at a price the company already chose. Buyers compete for allocation, not for price. The book closes before the question is ever asked.
Look at what sits inside the queue.
Elon Musk keeps 82.4 percent of the votes after the sale, with founder shares that vest on a Mars colony and 100 terawatts of compute in orbit, milestones no shareholder can enforce. Starlink is real: $11.4 billion in revenue, a 63 percent margin, 10.3 million subscribers across 164 countries. But the same building houses an xAI that lost $6.4 billion in a single year, stacked on a Starship capex furnace that eats the cash engine many times over. The compute contract paying $1.25 billion a month can be cancelled in 90 days. The first public earnings report lands in November.
Elizabeth Warren asked the SEC to stop it. Jim Chanos is short and calls it hopes and dreams. Morningstar prices it at $780 billion, less than half. The queue does not care.
What the queue is buying is the next ninety days of forced flow. What the queue is not buying is what the public shareholder actually owns. Two different assets. Two different numbers. Not even close.
It is the cleanest forced-flow infrastructure listing ever executed, and the first true public market test of a brand new asset class.
Tonight they price the queue. The piece prices the gap.

Today the people who hold $120 billion of crypto cannot agree on whether they just became prey.
That disagreement is the trade.
The model is Claude Fable 5, the first public release in Anthropic’s new Mythos tier. The restricted version of that tier has already surfaced more than 10,000 critical vulnerabilities across the world’s most important software. The public version routes anything that smells like offensive hacking back to the older model, and Anthropic says more than 1,000 hours of bug-bounty attacks failed to find a single universal jailbreak.
The crypto industry split in half within hours.
One camp, led by a Moonrock Capital founder relaying a top white-hat, says the cost and skill to find an exploitable flaw in a smart contract is about to fall to basically zero. Their advice was not analysis. It was a fire drill. Revoke your approvals. Pull funds from anything unaudited. Move to fresh hardware wallets. Treat every contract as a sitting duck until proven otherwise.
The other camp, led by the founder of Curve Finance, says calm down. His point is sharp. The model is brilliant at browser and operating-system bugs, but most smart contracts are small, already heavily audited, and easier for existing tools to read. The real money was never lost in the Solidity. It was lost at the edges. Compromised multisig keys. Poisoned frontend dependencies. A single cross-chain verifier with no backup.
Both of them are right, and that is the uncomfortable part.
The same machine that lets a defender audit a protocol in an afternoon lets an attacker probe it in the same afternoon. The advantage does not go to the side with the better technology anymore. It goes to the side that points it first. For years the asymmetry in crypto favored the attacker because defense was slow and manual. Frontier AI does not erase that asymmetry. It accelerates both ends of it at once and dares each side to move faster than the other.
Here is what the panic misses and the calm misses too.
The exposure was never really the code. It was the belief that nobody capable enough was looking. That belief is what just died. $120 billion sat behind the quiet assumption that the cost of finding the hole was higher than the value of the hole. Drop that cost toward zero and every unaudited contract is repriced in an instant, whether or not a single one is ever drained.
The machine did not create the risk. It revealed the price of pretending the risk was not there.
The protocols that survive the next year will be the ones already running this tool against their own code today. The ones that wait to find out which camp was right will not get to choose.