If There Is a Stock Market Crash, Here Are 2 Artificial Intelligence (AI) Stocks I Am Loading Up on

The article highlights Space Exploration Technologies (SpaceX) and Intel as potential long-term beneficiaries of the AI boom, but warns both look overvalued at current prices (SpaceX market cap $1.9T vs $7.8B latest-quarter revenue; price-to-sales ~65; Intel trading at 72.5x forward earnings vs 20.9x IT peers). SpaceX posted Q2 revenue up 92% YoY to $7.8B and AI revenue up 247.5% YoY to $2.6B, while Intel Q2 revenue rose 25% YoY to $16.1B and adjusted EPS was $0.42 vs -$0.10 a year earlier. Net: despite strong operating momentum, the pieces’ main takeaway is to wait for a downturn to potentially buy at better prices.

Analysis

The market is likely over-assigning software-like multiples to businesses whose cash flows are still lumpy and execution-dependent. For SpaceX, the biggest second-order risk is that the AI-compute line is being valued as recurring infrastructure demand when it may behave more like cyclical capacity monetization; in a risk-off tape, that distinction matters because late-stage private marks and public space proxies can compress faster than fundamentals deteriorate. The cleaner public expression of this thesis is not the name itself but the broader space/launch basket, where funding sensitivity and duration risk are highest.

Intel’s setup is better operationally, but the bar is much lower than the stock price implies. If AI agents really increase CPU intensity, that helps server demand, yet it does not automatically justify a premium multiple while manufacturing credibility and competitive share remain unresolved. The market may be missing that the first beneficiaries of rising AI capex are still the entrenched GPU ecosystem and hyperscale enablers; Intel needs a visible step-up in gross margin and share gains to defend this re-rating.

Near term, the signal is mostly valuation rather than a fundamental break. Over the next 1-3 months, any broad equity drawdown should be treated as a better entry point only if it coincides with verifiable operating momentum: higher server CPU share for Intel or externally measurable launch-cost improvement for SpaceX. The contrarian risk is that the move is overdone on the downside if AI demand broadens faster than expected, but absent that, current prices still look like they discount too much perfection.

Key Decisions for Investors

• Do not chase INTC here; set a conditional buy alert only on a 15-20% pullback or after forward P/E compresses materially, with the thesis invalidated if next two quarters do not show sustained server-CPU share gains and gross margin expansion.
• If we want AI exposure, prefer NVDA or GOOGL over INTC on a relative basis: long NVDA/GOOGL, short INTC only if upcoming guidance confirms that CPU demand is not translating into durable operating leverage at Intel.
• Fade public space proxies on market-wide risk-off days: consider short-dated put spreads on RKLB or ASTS rather than trying to express a view in SPCX, which is too valuation-sensitive to private-mark noise.
• Treat any SpaceX weakness as a watchlist opportunity, not a standing long: require evidence that AI revenue is recurring and margin-accretive before paying a premium multiple; otherwise wait for a broader liquidity-driven reset.
• Falsifiers to monitor: Intel raising full-year margin guidance and server-share commentary that beats by a wide margin; for space, a step-change in launch cadence or measurable cost-per-kilo decline would argue the multiple is less stretched than it appears.