Why is Meta's new gadget so cute? What to expect from Muse, which many are already calling “Tamagotchi”

In recent months, the major players in the field of artificial intelligence have accelerated the development of so-called AI agents — systems designed not only to respond to user requests, but also to autonomously carry out a range of operations. OpenAI has introduced an agent on ChatGPT capable of completing tasks of varying complexity; Anthropic is doing the same, while Google is integrating similar features into Gemini. Microsoft is also incorporating agents into its own applications, following a direction shared by numerous other startups in the sector. And recently, Meta has taken a significant step forward in this direction as well.

At Meta Connect, one of the main annual events organized by Mark Zuckerberg's company, a new artificial intelligence system was unveiled, designed for use with the new Meta VR Glasses and a wearable device meant to be worn around the neck. The latter calls to mind the increasingly common corporate mascots used by big tech companies to make these services feel less cold and more accessible to users.

Muse stands apart from traditional chatbots by performing a broader range of tasks. Interaction with the service takes place primarily through a text field, where users can enter their requests as they would with any chatbot — but the system doesn't stop at providing answers and can directly carry out certain actions.

How Muse Works

Through a browser built into the service, Muse can autonomously browse the web and interact with the apps and portals that users rely on daily — such as email or calendar, among many others. The system can also operate when not being directly used, for instance by monitoring the price of a given service at regular intervals and purchasing — with the user's consent — the most affordable option. Muse's basic services are free, but there are also plans at $20 and $100 per month that unlock additional features.

Other companies in the sector, including OpenAI and Anthropic, had already introduced agents capable of autonomously performing specific tasks online. With Muse, however, Meta appears to be offering the general public — not just tech-savvy users — a broader and more versatile system, designed to handle a diverse range of activities.

What Works and What Doesn't About Muse

Muse has generated considerable interest among industry insiders and technology experts — the New York Times, for instance, noted that the system's capabilities and autonomy are quite remarkable. For now, the service is available only in the United States and Canada through a dedicated app, but Meta is working to bring it to Europe as well — where its rollout may take longer, in part because the company will need to demonstrate that the system complies with Europe's strict privacy and personal data protection regulations.

Muse has already raised some concerns about user safety, at a time when there is much debate about the risks associated with the autonomy of AI agents — which have already shown, in several cases, the ability to carry out actions and make decisions that even their own developers had not anticipated. Meta maintains, however, that Muse was designed with these potential risks in mind, limiting the system's access to consumers' most sensitive and personal information.

It is partly for this reason that Amazon blocked access to Muse just days after its launch, arguing that using an AI agent to make purchases on its platform violates the site's terms of use; other players in the online shopping space, such as Shopify, have instead chosen to partner with Meta to integrate Muse as seamlessly as possible.