The setup is a dip-buying argument: oil back above $100, the 10-year at 5%, a likely Fed hike tomorrow at 2:00, and AI slowdown headlines have knocked down names the notes frame as structurally tied to AI capex. The thesis is that the slowdown talk is mostly coordination/regulatory capture, not real spending compression, because Anthropic, OpenAI, META, AMZN, ORCL and CRWV all still have massive contracts, backlogs, and buildouts. The video says the best trades are the beaten-down AI infrastructure names — QCOM, ORCL, APLD, and CRWV — while warning that a harder-than-expected Fed or more hawkish guidance could push every one of them lower first. Sponsored material is a separate VWAV pitch for AI drones and defense systems with multiple pending acquisitions, but it carries heavy dilution and execution risk.
Oil back above $100, the 10-year at 5%, a likely Fed hike tomorrow, and AI slowdown headlines are the stated reasons the market is down.
In an average year the market drops 14% at some point, and 3 out of 4 of those years still finish green.
NVDA, META, and Palanteer each lost 2/3 or more in 2022, and buyers of those dips got rich.
Anthropic's CEO wrote the AI slowdown essay, but Anthropic signed a $35 billion compute deal two weeks earlier and reportedly committed over $275 billion to compute this year alone.
OpenAI paused Astra last month and shipped it 3 weeks later; Elon Musk signed the pause-AI letter in 2023 and then started his own AI company four months after.
The essay asks for outside inspectors in every lab, coordination between the big companies, and a crackdown on the cheap tricks smaller companies use to catch up; David Sax called it regulatory capture.
AI spending is framed as unable to slow: last year data centers were about 4% of the economy but 92% of its growth, and Gartner just raised its forecast for next year.
QCOM is 31% off its high because phone sales were soft and Apple is using less of its stuff, but AMZN signed a deal last week for QCOM custom AI chips, put $4 billion behind it, and put up to $60 billion of orders on the table.
QCOM's revenue from the AMZN deal starts this December, and the stock trades at 19x earnings with a 2% dividend versus sector leaders around 40x-50x or more.
ORCL is about 56% off its high despite reporting the best quarter in its history: cloud revenue more than doubled and guidance was raised.
ORCL now builds giant AI data centers and rents them to OpenAI, META, and NVDA, but fell because it is spending $28 billion a quarter to build the capacity and borrowing to do it.
ORCL's customers have already signed $638 billion of contracts, described as the biggest backlog in software history, and Larry Ellison canceled a planned $7.5 billion stock sale and sold zero shares.
APLD is the AI-buildout landlord: it builds buildings, power and cooling, signs 15-year leases, and sits down about half from its high while burning cash and carrying debt.
$36 billion of signed leases sit against APLD's $7 billion market cap, most turning on in 2027 and 2028, and the next earnings date is October 8th.
CRWV rents AI compute by buying NVDA chips by the tens of thousands and leasing them to OpenAI, META, MSFT, and Anthropic; revenue more than doubled last quarter, but the stock is 56% off its high and down another 7% on Monday.
CRWV carries about $20 billion of debt, has $129 billion of signed contracts, including $22 billion from OpenAI and up to $35 billion from META through 2032, and Moody's rated the debt investment grade this year.
No contract cancellations have been seen despite the slowdown essay selloff, and a harder-than-expected Fed tomorrow at 2:00 is the main stated trigger for more downside across every stock mentioned.
VWAV, the sponsor, is a NASDAQ-listed Vision Wave Holdings pitch building AI drones and defense systems, with a July 2025 IPO, a $100 million shelf filed in August, and a convertible note issued last week.
VWAV says substantially all technology is still in development, has not generated meaningful revenue, and most deals are pending while most subsidiaries operate in Israel during an active conflict, making dilution and execution the main risks.