From Business Insider · Georgia Hennessy · · 1m 40s
Meta's Muse hits No. 1 app, but AI agents still stumble
Meta's personal AI agent Muse topped the App Store in its first week, yet early users reveal both stunning wins and glaring failures.
In brief
Meta's personal AI agent Muse topped the App Store in its first week, yet early users reveal both stunning wins and glaring failures. AI agents can handle complex real-world tasks like finding insurance deals, but they still struggle with basic capabilities and face structural limits from retailers protecting their…
Meta's Muse hits No. 1 on App Store
Meta's new personal AI agent Muse climbed to the top of Apple's free apps chart in the US, beating ChatGPT just one week after launch.
“a week after launch, muse is now the #1 app in the App Store!”
What makes Muse different from chatbots
Unlike chatbots that answer questions, Muse is an AI agent designed to act on your behalf. It can research, fill forms, buy items, book reservations, and connect to services like Spotify and OpenTable.
Insurance savings in five minutes
Entrepreneur Joe Devoy uploaded his auto insurance policy to Muse and asked for better rates. The agent found a policy saving him $3,500 a year, bought it, and canceled his old one in about five minutes.
“it found a policy saving me $3,500 /year, bought it, and canceled my old one.”
Shopping and recipes, handled instantly
Early users reported Muse finding a green puffer jacket with live discount codes, applying them, and buying at the lowest price. Another user said it found a recipe and loaded ingredients into a Whole Foods cart.
“find live discount codes, add to cart, apply, buy at cheapest price”
But Muse still has real limits
One user found Muse could only see the first frame of an Instagram video, even though Meta owns Instagram. Agents also struggle with accurate pricing data and face resistance from retailers protecting their business models.
“it only could see the first frame so asked me to share the link”
Privacy and accuracy remain open questions
As AI agents go mainstream, concerns linger about potential errors and the privacy trade-offs of letting them handle sensitive tasks like insurance and banking.
What it means
AI agents can handle complex real-world tasks like finding insurance deals, but they still struggle with basic capabilities and face structural limits from retailers protecting their business models.
