Michael Burry Just Said 7 Words Dismissing AI After Slowdown Fears: "There Is Nothing AI to Slow Down." Should Investors Trust His Nvidia and Palantir Shorts?

Michael Burry closed his Nvidia put position but retained Palantir puts and other technology shorts, arguing that calls by AI leaders to slow development are self-serving rather than evidence of genuine AI-risk concerns. Nvidia is up 13% year to date and trades at 27x earnings, while Palantir is roughly flat year to date but trades at a 148x P/E. The article highlights risks to AI valuations if development momentum or perceived technological value weakens, while noting forecasts for $3 trillion-$4 trillion of AI-related capital expenditure by 2030.

Analysis

The investable signal is not the commentator’s positioning; it is whether AI-safety rhetoric turns into procurement friction or regulation. A voluntary pacing narrative is more likely to raise compliance costs and entrench scaled platforms than to reduce near-term accelerator demand: MSFT, GOOGL and AMZN can absorb safety, audit and compute requirements that smaller model developers cannot. That is modestly supportive of NVDA’s customer concentration and pricing power over the next 1-3 quarters, while potentially squeezing smaller GPU renters and unprofitable application vendors.

PLTR remains the cleaner duration-risk expression. Its valuation requires sustained revenue growth, expanding commercial conversion and limited stock-based-compensation dilution; any evidence that enterprise AI pilots are extending rather than converting can compress the multiple sharply even if reported revenue remains healthy. Conversely, PLTR’s governance, security and deployment tooling could benefit if regulated customers require auditable AI workflows, making a simple outright short vulnerable to contract announcements or raised guidance.

The overlooked distinction is that a slowdown in frontier-model capability does not necessarily mean a slowdown in AI monetization. If model progress plateaus, capital allocation should shift from training clusters toward inference optimization, networking and enterprise integration—relative beneficiaries include AVGO and Arista Networks (ANET), while the highest-multiple application names lose scarcity value. This article alone is low-impact; the decisive catalysts are hyperscaler capex guidance, NVDA’s supply/lead-time commentary, and PLTR’s commercial revenue and remaining-deal-value trajectory over the next two earnings cycles.

Key Decisions for Investors

• No directional trade solely on the reported investor view; treat it as a sentiment alert. Reassess after the next MSFT/GOOGL/AMZN capex updates: a second consecutive aggregate capex-guide cut would be the actionable confirmation of a semiconductor demand rollover.
• Express valuation dispersion over 3-6 months with long NVDA / short PLTR in roughly beta-neutral dollar amounts. NVDA has nearer-term earnings support from installed demand, while PLTR carries greater multiple sensitivity to commercial-conversion misses; exit if PLTR raises full-year commercial-growth guidance materially or NVDA cuts data-center outlook.
• For a less binary AI-spend rotation, accumulate AVGO or ANET on broad AI volatility rather than chase application software. The thesis is a 6-18 month migration toward inference, networking and custom silicon; invalidate if hyperscaler capex falls materially rather than merely reallocating from training.
• If maintaining PLTR downside exposure, prefer defined-risk put spreads 3-6 months out rather than naked puts. The relevant downside catalyst is a miss in net-new commercial bookings or a deceleration in remaining deal value, not generalized AI-safety headlines.