Peter Yang Tests Muse: Bill Discounts, Marketplace, Habits, and Phone Calls

Peter Yang demonstrates Meta’s Muse personal agent across bills, messages, shopping, a news feed, habits and local activities. He calls it his favorite personal agent at the time, a subjective assessment based on his own use.

For his AT&T phone account, Muse identifies an autopay discount for switching from a credit card to a bank account. Yang says he enters the bank and routing details himself and enables the change. He reports a $40 monthly reduction across four lines, or about $480 per year if the discount continues.

Yang customizes Muse’s avatar into an animated red panda. He can select a generated appearance, but his attempt to add additional custom animations to the profile runs into a limitation. He likes the personalization despite that failure.

Muse checks messaging services for items needing follow-up and asks to connect other accounts. Yang says his Instagram and Threads messages mostly contain spam in this example. He likes a scheduled digest for multiple messaging apps, though his attempt to connect Linear does not work.

For Facebook Marketplace, he asks the agent to find nearby golf clubs and describes how it could help list his old phone. He also demonstrates asking it to contact a seller, with a confirmation before sending. Search results, availability and any transaction terms still require review.

Yang asks for a personalized morning feed covering AI news and his projects. He likes being able to skim it and open stories for more detail. He describes the feed as ad-free in his test; future advertising is his speculation, not a product announcement.

To change a bedtime habit, Yang asks for help leaving his phone outside the bedroom. Muse sets reminders and builds a tracker for daily check-ins and streaks. Yang reports that his first attempts have not gone well, so the display is a record of an ongoing experiment, not proof the habit has changed.

An Ideas tab proposes tasks based on his conversations. Yang tries a side-by-side electric-shaver comparison; he finds the retailer-based list a useful start but wishes it had consulted independent review sites. He also browses suggestions for local activities and planning.

For an anniversary outing, the agent finds events and a nearby dinner option. Yang discusses asking it to book tickets but does not show a completed purchase in this example. He also suggests a recurring weekend scout for local events.

Yang then has Muse call Xfinity to ask about his internet bill. The agent identifies itself as calling on his behalf and receives information about a possible promotional price and autopay discount. The transcript says Yang must call back and authenticate before any change is made.

Although Yang opens with a claim of more than $800 in annual phone and cable savings, the shown evidence supports an activated $480 annualized phone discount and a separate potential cable offer requiring his follow-up. The actual total depends on the offer being accepted and remaining in effect.

Yang concludes that Muse is strongest for his personal tasks and describes using other tools for work. He speculates about Meta’s distribution and stock prospects; these are his opinions, not evidence of future adoption or investment returns. Calling access in his case followed a request to the product team, so availability may differ.