Meta Platforms' recently launched personal AI agent, Muse, is proving to be a significant success, with early adoption rates surpassing those of rival AI assistants. Within its first two weeks, Muse garnered more than 2.5 million global downloads, outperforming Anthropic's Claude and Google's Gemini. Notably, it has also surpassed OpenAI's ChatGPT in terms of iOS mobile downloads, according to data from Jefferies, Sensor Tower, and UBS. This rapid uptake has led JPMorgan analysts to suggest that Muse could become the most widely used consumer AI application since ChatGPT, while Jefferies has raised its price target for Meta to $875 from $710, indicating an 18% potential upside.
This success is refocusing Wall Street's attention on consumer-oriented AI agents, moving away from the previous emphasis on enterprise AI applications. Analysts from Evercore ISI and UBS believe that consumer AI agents could drive the next phase of the AI trade, offering new monetization opportunities beyond business applications and stimulating demand for computing power. The market for digital consumer spending that could be captured by such agentic products is estimated to be massive and multifaceted, with Morgan Stanley analysts seeing roughly $30 trillion in consumer spend that could be digitized across various sectors like online advertising, e-commerce, travel, and food delivery.
The rapid rise of Muse is not without its challenges and implications. Amazon has reportedly blocked Muse from its website, preventing it from acting as a personal shopper for users, stating that third-party applications should operate openly and respect service provider decisions. Additionally, analysts at Deutsche Bank and Bryan Kraft suggest that Muse could lead to increased app cancellations and software churn, particularly for streaming services and news subscriptions, as it could help users manage and optimize their subscriptions. This success also brings into question Meta's nascent cloud business, as the high demand for Muse may lead Meta to retain its excess computing capacity rather than renting it out.
