SpaceX just punched through $160 and is sitting around a $2.2T market cap. Highest print since early July. The move that actually matters to me is the rename. Elon is dropping SpaceXAI for SpaceXSI and calling the company a super intelligence shop, not just a launch company. Rockets were the wedge. Starlink is the cash engine. Now the pitch is compute and intelligence in orbit. If that framing sticks, this stops trading like a space contractor and starts trading like the other AI platforms. Still wild that a company that was “too risky for public markets” a few years ago is now one of the biggest names on the board. Other things moving today, not the same story: the SEC approved the first US 3x leveraged Bitcoin and Ethereum ETFs (futures-based, via Volatility Shares on Cboe), and $NVDA is being credited with dragging the QQQ to new highs, with one note that it now does about $8M in revenue per employee.

Don't forget about Optimus, he was also there!

Been staring at this SpaceX stat longer than I should. They now own all 5 fastest U.S. launch-to-docking times to the ISS. Crew-13 did it in 7 hours 55 minutes on Oct 1. The next four are Dragon too. People still treat the rockets like a side quest. The cadence is the product. Same week Elon cut RAM on the Tesla AI5 and AI6 chips because there isn’t enough memory volume for Optimus. Bandwidth stayed flat, storage got cut, and he basically shrugged. That’s the part I keep coming back to. Rockets on one side, HBM on the other. $MU, SK hynix, the whole memory stack is the constraint, not the slide deck. $NVDA can ship the accelerator. Somebody still has to feed it. Not a call. Just what I’m watching while everyone argues about whether AI is a bubble.

SpaceX just made the trip to the ISS feel almost routine. Crew-13’s Dragon launched and docked in 7 hours and 55 minutes, the fastest U.S. flight from the pad to the station. Eight hours from Florida to a spacecraft a few hundred kilometers up. A few years ago that was a multi-day mission. Now it’s closer to a long commute. This is the part people still underrate. Reusability wasn’t the end state, it was the setup. Faster turnarounds, tighter windows, and a cadence that starts to look like infrastructure instead of a stunt. Starlink already runs on that logic. Crew and cargo are catching up. If you’re watching the space trade, this is the signal, not the headline. The bottleneck is no longer “can we get there.” It’s how often, how cheap, and what you build once the trip is boring. $SPCX

United just said 880+ planes get Starlink by the end of 2026, and the whole fleet by the end of 2027. They’re already live on 560+ aircraft. 14.2 million devices connected across 464,000+ flights. That’s not a pilot program anymore. That’s in-flight wifi turning into a real network, and it’s SpaceX selling connectivity the way carriers used to sell seats. If Starlink keeps locking airlines like this, the IPO story stops being about rockets and starts being about a recurring revenue machine sitting on top of the constellation. $SPCX

Been staring at this Micron note and it still doesn't sit right. Morgan Stanley is basically saying $MU earns close to half its current market cap by the end of 2028, and most of that gets handed back to shareholders. Half a trillion in earnings over roughly two years. For scale, $AMD is expected to do about $3B next quarter. Memory is no longer the boring cycle people keep calling it. HBM is the bottleneck, and Micron is one of three names that actually ships it. If that number is even directionally right, the multiple on $MU is still pricing a downcycle that already ended. $NVDA, $AMD, $MRVL all need the same wafers. The picks-and-shovels trade just got a lot louder. $MU

Might actuallynhave to use them on my next flight from HKG to BKK

FAA just opened the environmental review for Starship out of Cape Canaveral. Up to 76 launches a year from SLC-37. 152 landings. Boosters coming back to the pad, ships too. SpaceX is already pouring concrete while the paperwork catches up. Comments close October 14. Starbase was the prototype factory. Florida is how this becomes a cadence. If they actually get anywhere near 76 a year, the launch bottleneck stops being a talking point and starts being a solved problem. Still wild that the same company building this is also the one selling AI compute by the gigawatt. $SPCX

Saylor just dropped “more orange than ever” on a Sunday afternoon. I’ve watched this long enough to know he doesn’t post that for vibes. Same line shows up right before Strategy adds to the stack. Not calling the size, but the pattern is pretty consistent. Monday open should be interesting. $BTC $MSTR

Or just SXSI for short

Elon just called SpaceX a super intelligence company. Not a rocket company. Not a satellite company. Read that sentence twice!

Everyone still prices SpaceX like a rocket company. Elon just called it a super intelligence company, and the tape already started agreeing. Friday was the tell. Crew-13 docked, SPCX ripped about 7%, Tesla printed 486k deliveries and finally looked like the EV winter was over, and NVDA and AMD both tagged new highs. The quiet part is the stack underneath. You do not run orbital compute, Starlink, or Tesla’s autonomy stack without HBM and networking. That is MU and SK hynix on memory, ALAB and MRVL on the interconnect, AMD as the second source Nvidia cannot fully crowd out. Rockets are the visible product. The business is selling power, bandwidth, and inference to everyone who cannot build a data center fast enough. If that framing sticks, the chip suppliers get paid either way. Not advice. Just how I’m reading the week.

Bitcoin just pushed back over $85k and I’m not treating it like noise. ETH is back above $2,700 and SOL is participating instead of lagging, which is usually the tell that this isn’t just a BTC squeeze. Weekly close is the only thing that matters from here. Hold $85k into Sunday and the “Uptober” crowd stops sounding delusional. Lose it and this was just another dead-cat bounce on thin weekend books. Still watching the AI complex underneath this. NVDA, AMD, MU and the rest of the memory/networking names have been the real bid all year. If crypto is catching a risk-on bid while those charts stay intact, that’s a different tape than the September scare. Not married to the $100k calls. Just want the weekly close. $BTC

I keep staring at Micron and it still doesn’t feel real. $MU is up more than 480% in a year. About a trillion dollars of market cap, just added. The stock is over $1,000. A dollar in it five years ago is roughly $15 now. This week they beat and guided above the Street. Everyone spent two years arguing about who sells the GPUs. The quiet trade was the memory those GPUs actually eat. HBM isn’t a side quest anymore. It’s the bottleneck. Not saying chase it here. Saying the AI buildout isn’t only an $NVDA story. $MU just proved the rest of the stack can reprice just as violently. $MU

Jason Oppenheim just handed his Bentley to his dad and bought a Model Y. Not a Tesla influencer. The Selling Sunset guy. Runs a real estate shop that has done $5B+ in sales. He posted what he called the most important video he’s ever made: hit the button in Newport Beach, didn’t touch the wheel for an hour and 15 minutes into LA, car parked itself. Then he called his brother, who bought one while they were still on the phone. Now he’s putting Teslas with FSD in front of 10 employees. That’s the part I keep coming back to. Luxury used to mean a driver and a badge. The flex is becoming “I don’t drive.” When someone who can buy anything picks the software over the Bentley, FSD stopped being a demo and started being a reason people switch. Still early. But this is how it actually spreads. Not another chart. A guy in Newport Beach telling his brother to just buy one. $TSLA

SpaceX and Tesla both ripped yesterday and Elon added about $61B in a day. Net worth is back knocking on $1T. Still weird to watch one person swing that much on two companies that barely look related on paper. Rockets on one side, cars and robotaxis on the other, and the market keeps pricing them like the same bet. I don’t think the tape is done with this. SpaceX is still the cleaner AI-infrastructure story if orbital compute is real, and Tesla is still the one that has to prove the delivery and autonomy numbers. Yesterday the market just decided both get the benefit of the doubt.

Do you remember when you joined X? I do! #MyXAnniversary