2 AI Stocks I Don't Like (Including Palantir) and 1 I Love

Palantir posted Q4 2025 revenue of about $1.41B, up 70% year over year, and Snowflake delivered product revenue of $1.23B, up 30%, but both are criticized as still too expensive relative to fundamentals. The article is more constructive on Amazon, which reported Q4 net sales of $213.4B (+14%) and AWS revenue of $35.6B (+24%), with AWS on a $142.4B run rate, though free cash flow fell to $11.2B in 2025 due to heavy AI-related capex. Overall tone is selective and valuation-focused, favoring Amazon over high-multiple AI software names.

Analysis

The market is beginning to separate monetizable AI infrastructure from AI narrative software. That is the key second-order effect: capital is likely to keep flowing toward platforms that can absorb AI demand with existing distribution and balance-sheet capacity, while premium software names face multiple compression if growth merely normalizes instead of accelerating. In that context, AMZN is the cleaner beneficiary because AI spend at scale reinforces a flywheel in cloud, logistics, and retail monetization rather than relying on a single valuation story.

PLTR and SNOW are vulnerable to a classic post-hype de-rating regime: even strong operating prints can fail to support current multiples if investors stop paying for duration. For PLTR, the risk is not business deterioration but expectation exhaustion — once growth decelerates from exceptional levels, the stock can underperform for several quarters even if fundamentals remain healthy. For SNOW, the market is effectively underwriting a future margin structure that is still unproven; until operating leverage is visible, every incremental revenue dollar risks being treated as low-quality growth rather than profit expansion.

The contrarian miss is that the recent selloff may not be a broad AI unwind, but a bifurcation within AI beneficiaries. That favors cash-rich infrastructure owners and punishes pure-play software with long-dated profitability. If the next two quarters show sustained AWS reacceleration and continued AI capex digestion, AMZN could rerate further even if near-term FCF remains pressured; meanwhile, PLTR/SNOW could stay range-bound or lower as sentiment resets.

Catalyst-wise, the next inflection points are quarterly guidance revisions and capex commentary, not backward-looking revenue prints. A reversal in PLTR/SNOW likely requires either a sharp multiple-supportive macro tape or proof of operating leverage; absent that, rallies are better faded than chased. The main tail risk to the bullish AMZN view is a prolonged capex digestion phase that delays FCF recovery by 2-4 quarters, but that is more a timing issue than a thesis breaker.

Key Decisions for Investors

• Go long AMZN on any 3-5% post-earnings pullback over the next 2-6 weeks; use a 12-month horizon targeting further multiple support from sustained AWS growth, with downside protected by durable cash generation.
• Initiate a relative-value pair: long AMZN / short SNOW for 1-3 months to express the market’s preference for profitable AI infrastructure over unprofitable growth; thesis breaks if SNOW demonstrates two consecutive quarters of operating leverage.
• Avoid initiating new longs in PLTR after sharp drawdowns unless you can underwrite another 1-2 quarters of >50% growth; the better risk/reward is waiting for a lower entry or using call spreads instead of stock.
• Consider a tactical short or put spread on SNOW into strength over the next earnings cycle; the setup favors multiple compression if guidance remains solid but profitability remains absent.
• For AI exposure, rotate out of pure-play software beta and into diversified mega-cap platforms such as AMZN; that reduces single-name execution risk while keeping exposure to AI capex and cloud demand.