The video argues that SpaceX’s earnings and AI/data-center execution materially strengthened the bull case, even though the stock was under pressure from an imminent unlock of roughly $100 billion of shares. The hosts frame SpaceX as a three-pillar business — connectivity, AI, and space — with Starlink growth, heavy capex, and a Jensen Huang/Nvidia partnership pointing to a much larger revenue path, including Elon Musk’s claim of $1 trillion in revenue by 2030. At the same time, they repeatedly flag the near-term risks: dilution, short interest, timeline risk, key-man risk, and the possibility that the stock sells off even if the long-term thesis improves. The ending stays constructive but cautious, saying the best setup may be to wait through the unlock volatility while keeping the long-term upside in view.
They said August 4 was the day SpaceX had to prove itself after Wall Street analysts were broadly bearish on the first earnings report, and they claimed SpaceX "absolutely smashed" it.
SpaceX revenue was said to be $7.81 billion, up 92% year over year, versus a $6.93 billion estimate, and Starlink subscribers doubled to 12 million.
The earnings surprise included EPS of 0.09 versus expectations for a 26-cent net loss, and net loss improved to about $500 million instead of the expected $2 billion loss.
The stock was said to still be down because of the market’s concern that about $100 billion of SpaceX stock would go live in an unlock the next day, with the shares later described as down about 9% to $115.
Josh said Elon Musk claimed on the earnings call that SpaceX will "build robots with mesh drivers on the moon" and "1000x our economy on Earth because of it," while also saying SpaceX expects $1 trillion in revenue by 2030, not 2031.
The speaker broke SpaceX into three pillars — connectivity, AI, and space — and said AI revenue was up 250% year over year while space revenue was up 30% year over year.
The speaker said SpaceX’s capex came in at just under $18.5 billion, 39% above expectations, which they noted was almost NASA’s entire $25 billion annual budget spent in a single quarter.
They said SpaceX already has roughly $50 billion of signed backlog and that CFO Bret Johnson argued not all capex is equal because SpaceX’s data-center spend pays back in a single calendar year.
The panel said SpaceX is not just an AI company but a computation company with the largest data-center cluster on Earth, the largest GPU footprint of Nvidia GPUs, and a major partnership with Jensen Huang.
They said Elon Musk told Jensen SpaceX would be exclusive to Nvidia, choosing the Vera Rubin architecture, and that Musk initially said 10 gigawatts by end-2027 before later guiding toward roughly 20 gigawatts by around 2030, with a floor of 15 gigawatts.
The speaker said 15 gigawatts of solar capacity by the end of the decade would imply about $600 billion to $750 billion of revenue, and that 10 gigawatts of GPUs by the end of next year could mean $200 billion in Nvidia revenue, or $100 billion if it reaches 6 gigawatts.
The video argued the AI bubble is really a GPU capex bubble, but that cloud revenue at Amazon, Google Cloud, Azure, and AWS is still growing 40% to 45% compounded each quarter with expanding margins, which is why the compute business looks profitable.
On Starlink, they said subscribers doubled year over year to 12 million, with 1.7 million added in the quarter, and that version 3 satellites could launch on the next Starship, with the constellation starting to launch as soon as next year.
Gwynne Shotwell was cited saying SpaceX wants to deploy something similar to Verizon or AT&T and become a major pipeline for global internet usage, including direct-to-cell service that could bring high-bandwidth internet to iPhones by this time next year.
The closing view was that SpaceX has a bright future but is highly risky in the short term: there is key-man risk around Elon Musk, timeline risk around 15 gigawatts by 2030, and 25% to 36% of the share supply unlocking over the next couple of months could create major volatility.
The hosts compared the setup to Tesla’s 2020-to-2021 squeeze, saying SpaceX could be a similar violent repricing if the thesis proves out, but that they may wait for the unlock-driven volatility before adding more.
They ended by stressing the manufacturing moat angle, saying SpaceX’s ability to deploy data centers quickly, pay back in 12 months, and move atoms at scale puts it alongside China and Tesla as one of the most capable manufacturing machines in the world.