Meta Conducts Human Caller Test for AI Assistant Muse
Meta tested human contractors for its Muse assistant after businesses hung up their phones upon realizing it was an artificial intelligence.
Meta conducted a test for the phone-calling feature of its AI assistant Muse, employing human contractors after companies terminated calls upon realizing they were speaking with an artificial intelligence.
Muse and the Phone Feature Test
Meta ran a testing process that involved human contractors taking part in certain calls for the telephone-calling function of its personal AI assistant, Muse.
Based on internal communications and statements from a Meta spokesperson, this step was taken due to the issue where called businesses hung up the phone upon realizing they were dealing with artificial intelligence on the other end.
Employee Warnings and Termination of the Test
Meta employees warned that the human-involved testing process in question could lead to the leakage of sensitive information or weaken security measures.
Shortly after announcing this test to employees, company management informed them that it was halting this practice for the time being.
Statements from Meta Spokesperson
Meta spokesperson Daniel Roberts stated that the test was conducted to gather feedback prior to the public launch.
Roberts noted that employee feedback was extremely positive and that features would only be rolled out when ready with appropriate explanations.
Launch of the Application and Integrations
Meta launched its personal AI assistant Muse, which runs on a cloud-based computer and houses a Sentinel agent, on September 8.
Ten days after its launch, Muse climbed to the top spot on the Apple App Store rankings, while integrating Stripe's Link, Shopify's Shop Pay, and PayPal among its payment methods.