Meta Ray-Ban Display Glasses Malfunctions

Mark Zuckerberg’s much-anticipated demo of the $799 Meta Ray-Ban Display glasses stumbled during the Connect 2025 keynote in California. Two live showcases malfunctioned in front of a packed audience, raising questions about the readiness of Meta’s new wearable AI device.

Cooking Demo Goes Wrong

The first failure occurred during a collaboration with cooking creator Jack Mancuso. The glasses’ LiveAI feature was supposed to guide users through recipes step by step. Instead, the assistant repeatedly skipped instructions and refused to answer simple prompts. Tech Radar’s Lance Ulanoff, who attended the event, said the AI looked “confused and jumping around” due to Wi-Fi strain from thousands of attendees.

Neural Band Struggles on Stage

The second mishap involved the Meta Neural Band, the wrist-worn controller that allows gesture-based input. Zuckerberg successfully exchanged texts with Meta CTO Andrew Bosworth. However, when Bosworth tried to initiate a WhatsApp video call, the glasses failed to register. Zuckerberg eventually gave up as Bosworth walked onstage, joking about “brutal Wi-Fi.”

Industry Experts React

Observers noted that other tech firms, such as Apple, avoid live demos at major events to prevent such failures. “They use pre-recorded videos for total control,” Ulanoff said. He added that Zuckerberg’s risk-taking reflected his enthusiasm, even though it resulted in uncomfortable moments on stage.

Meta’s Response to the Setbacks

Despite the glitches, Mark Zuckerberg insisted that the glasses represent a leap toward Meta’s vision of “agentic AI.” The device is designed to anticipate user needs, handle everyday tasks, and integrate seamlessly with personal assistants. Zuckerberg acknowledged the mishaps with humor, telling the crowd, “You practice these things like a hundred times, and then you never know what’s gonna happen.”

The Road Ahead

The Meta Ray-Ban Display glasses, which go on sale September 30 in the U.S., are still positioned as a milestone in wearable AI. However, the failed demo highlighted the challenges of bringing experimental features to real-world settings. For Meta, the path to mainstream adoption may depend on proving that the technology works smoothly outside the keynote stage.