Meta Unveils Muse Wonder AI Models With Active Memory for Building Personal Superintelligence

Meta has unveiled Muse Wonder, a new family of language models designed to efficiently handle complex professional and everyday tasks. The developers describe the release as a major step toward creating personal superintelligence—artificial intelligence capable of deeply adapting to an individual’s habits and goals, retaining the context of interactions, and working autonomously toward the final result.

The lineup includes two versions: the base Muse Wonder and its more advanced variant, Muse Wonder Full. The smaller model has already been integrated into the Meta Model API and Muse Code, while the expanded version is available to Muse Code subscribers. In the near future, the family will be integrated directly into the consumer-facing Muse service, where it will handle the generation of complex materials and the execution of labor-intensive analytical workflows.

The key technological breakthrough of the new lineup is the Active Memory architecture, which enables the algorithms to reliably retain essential information about the user and retrieve it instantly when handling current tasks. The second major component is Active Focus technology, which coordinates sustained work on complex projects. It allows the AI to independently test alternative approaches, flexibly overcome systemic obstacles as they arise, and quickly incorporate new input, while in cases of critical ambiguity, it can ask the user targeted questions or seek approval for key steps.

In addition, the developers have significantly improved the models’ programming skills and ability to work with highly specialized information in law, finance, and medicine. The models were trained using long-term, complex scenarios requiring step-by-step quality control and the correction of their own mistakes, enabling them to achieve strong results in specialized benchmarks and when tackling fundamentally new tasks.

Because the concept of a personal assistant involves deep integration into an individual’s personal environment, including access to messages, calendars, and confidential documents, Meta has devoted unprecedented attention to security. The Muse Wonder architecture is designed to reliably protect personal data, successfully resist malicious external instructions, and block any attempts to use the AI in dangerous or destructive ways.