Muse review: The personal AI agent that gets consumer UX right

The episode begins with the host testing Meta's new personal AI agent, Muse. She shares her enthusiasm for its design, usability, and how it differentiates itself with a consumer-friendly approach. She highlights Muse’s capabilities including onboarding, goal setting, a family newsletter, browsing, shopping, and the animated avatar. Overall, she finds Muse to be the best-designed personal agent she has encountered, praising its delightful interface, trust-building permission system, and human-like tone.

The host explores key features of Muse, such as onboarding, setting up personalized identities, and activity tracking. She emphasizes the agent’s aesthetic appeal, the activity lineage for transparency, and the thoughtful permission prompts. She details her user experience, including personalized calendar management, PDF newsletters, local news, and the 'Ideas' prompt library. She also discusses setting and tracking personal goals (water intake, sleep training) and the innovative avatar customization, praising how these design choices contribute to a human-first, trustful interaction.

The host tests Muse’s browsing during shopping for shoes and movie tickets. While the shoe shopping experience isn't as successful, buying tickets for a movie (The Odyssey) is smooth, showcasing Muse's strengths in transactional tasks. She notes the importance of the integrated Stripe link for payments. She also observes some shortcomings, like incorrect search results for browsing, indicating room for improvement in internet browsing and transaction reliability. Overall, she appreciates Muse's design for everyday personal and family use.

Throughout the episode, the host highlights Muse’s entertaining design features, such as animated avatars that change and perform tasks, the activity feed with step-by-step task lineage, and media generation like creating podcasts. These elements make Muse feel personable and engaging. She demonstrates its ability to generate media content, such as a custom podcast about AI news, which she finds surprisingly effective and high-quality.

The host concludes her review by emphasizing Muse’s consumer-oriented design, trustworthiness, and aesthetic appeal. She plans to use Muse for wellness management, family organization, and local errands. She also considers exploring Facebook’s AI models due to the impressive artifact generation. She reflects on the importance of good product design in AI agents, noting Muse as a benchmark and encouraging product teams to learn from its strengths. The episode wraps up with a positive outlook on Muse's potential for everyday personal use.