In 2013, Facebook bet on a phone that would put the social network at the center of the screen. But the price of the HTC First plunged from $99 to 99 cents within weeks of its launch. Years later, Mark Zuckerberg made a much bigger bet on virtual reality and the metaverse, while operating losses at Reality Labs topped $80 billion. Now there’s a new wager: Muse Charm, a small device that gives users access to the Muse AI assistant. Has Zuckerberg really changed his bet—or just changed the shape of the device?

The device isn’t the whole idea

Charm doesn’t perform tasks on its own; it’s a way to access Muse, the assistant that’s supposed to search, organize, book, and shop on users’ behalf. Meta’s bigger bet is that the assistant will become the starting point for people’s requests, instead of having them open and switch between different apps and websites.

That’s where the economic opportunity lies. Meta today relies mainly on ads shown in its apps. Muse could add revenue from paid subscriptions or commissions on purchases and bookings it handles. If the assistant directs users to a product and completes the purchase, Meta could take on a new role in the relationship between retailers and customers. But this revenue is only potential; it’s not enough for users to download the app—they have to return to it, trust it, and pay for its services.

Why did stocks move?

The irony is that market moves were tied to Muse’s popularity before Charm was unveiled. Shopify shares rose after the company partnered with Meta to enable purchases through the assistant, while shares of companies such as Booking Holdings and Intuit fell amid investor concerns that the assistant could change how consumers search for trips, services, and products.

These moves reflect market expectations, not confirmed losses of customers for those companies. The assistant could open a new sales channel for its partners, while also shortening the path consumers take through other platforms. It’s not yet clear which side will secure the biggest share of the customer relationship: the assistant, the retailer, or the platform that provides the service.

The real test: usefulness and trust

Meta’s history helps explain investors’ caution. Past losses remind them that an ambitious idea doesn’t guarantee broad demand. Charm may remain a product for tech enthusiasts if it doesn’t offer benefits that a phone or smartwatch can’t.

The audio glasses Meta introduced point to another part of the test: removing the camera may ease some concerns about being filmed, but it doesn’t resolve privacy questions if the device records audio. For products that come close to everyday life, trust is part of the product’s value. Users won’t let an assistant read their messages or make purchases unless they feel its use is safe and transparent.

For Saudi retailers, the question worth watching is whether the spread of AI agents could eventually become another channel for reaching customers—or make the assistant the interface through which purchases are made. There’s no local data yet to settle the question, but Meta’s experiment shows what to watch: will ease of use lead to repeat visits, actual payments, and sufficient trust?

Zuckerberg’s test this time may not be selling a small device. It’s about making Muse an intermediary people rely on, then turning that reliance into revenue. Only then will it be clear whether Meta has found a new platform—or added another device to its list of bets.