Cathie Wood Just Made a $55 Million Bet That Google’s Search Box Is a Relic
(SeaPRwire) – By: Christian Pierce
ARK sold 84,392 shares of Alphabet worth $28.6 million on September 9, 2026. Then, on the same trading day, Cathie Wood’s fund bought 43,091 shares of Meta Platforms for $26.4 million. The timing is not accidental. Meta had just unveiled Muse, a new AI agent built to handle shopping, travel bookings, and emails. The stock jumped 6.55% that session. Alphabet fell 2.28%. Zuckerberg described Muse as part of a push toward “personal superintelligence.” That framing matters more than it sounds. It signals that Meta sees its AI spend as a replacement for the search interface, not just an addition to it. The company expects to spend between $130 billion and $145 billion this year. A large portion goes toward AI development. For Alphabet shareholders, that number should feel like a threat multiplier. Google has its own AI models. It has Gemini. But the question is whether Gemini can match the distribution advantage of a platform where billions of people already manage their social lives, messaging, and news feeds every single day. Distribution beats capability when the capability gap narrows. That is the uncomfortable truth this trade exposes. Mizuho analyst Lloyd Walmsley said Muse’s launch could renew concerns about Google’s Gemini AI models and the effect of AI agents on traditional search. That’s the anxiety driving capital out of Alphabet and into Meta. ARK has been trimming Alphabet positions for several weeks. Wednesday’s sale continues a deliberate de-risking pattern. It is not a single-day panic. The fund is making a structural bet that search monetization is decaying faster than Wall Street models assume.
On the biotech side, ARK bought 271,159 shares of Beam Therapeutics for about $7.3 million. The company’s stock fell 6.17% to $25.18 after releasing Phase 1/2 trial data for BEAM-302, a treatment for alpha-1 antitrypsin deficiency. ARK bought 47,113 shares of CRISPR Therapeutics for roughly $2.6 million. Those shares dropped 2.43% to $53.45. ARK also picked up 204,206 shares of Intellia Therapeutics for about $2.6 million. Intellia shares fell 4.16% to $12.21. The FDA recently accepted Intellia’s application for lonvo-z with Priority Review. A decision is expected by March 10, 2027. If approved, lonvo-z could be the first one-time treatment for hereditary angioedema using in vivo CRISPR gene editing. ARK trimmed Tempus AI by selling 22,767 shares worth about $1.5 million. Tempus shares fell 4.68% during the session. The fund has been selling Tempus across several recent trading days. That’s a rotation out of AI-adjacent diagnostic platforms and into gene-editing companies with nearer-term FDA catalysts. The logic is simple. Tempus AI depends on hospitals and insurers adopting its diagnostic layer. That adoption curve is slow and politically complicated. CRISPR companies have binary events. A single FDA approval can transform a company’s entire valuation overnight. Cash is moving toward companies with defined regulatory catalysts and away from companies with adoption-dependent growth stories.
The commercial loop here is straightforward. Meta owns the social graph. It owns the daily attention surface. Muse plugs AI task completion directly into that surface. Alphabet owns the query-response pipeline. That pipeline is 20 years old. It was designed for a world where the user types a question and expects a list of links. If agents can answer multi-step tasks without a search box, the query pipeline loses relevance. ARK’s portfolio adjustment reflects a conviction that this transition is accelerating. The shift is not about AI capability. It is about which company gets to be the interface between humans and autonomous action. Meta is betting that the interface is already in people’s pockets. Google is betting it is still in the browser tab. The practical implication for investors is concrete. If your valuation of Alphabet assumes search ad revenue stays flat through 2028, your position sizing is built on a distribution model that may not exist in three years. Re-assess before the next earnings report, not after the consensus breaks. If you hold Meta, watch whether Muse’s task completion rates actually translate into measurable engagement gains in the next two quarters. If they do not, the $145 billion spend becomes a capital destruction story. The money is moving. The question is whether it is moving toward the right interface. The answer will not be known for 12 to 18 months. But the positioning you make today will determine what you earn when it is.
Author bio: Christian Pierce, a chief financial columnist and markets commentator who has covered technology capital flows, AI investment strategies, and large-cap rotation dynamics for major financial publications across two decades.