Meta encountered significant technical difficulties during the live demonstrations of its new smart glasses this week. Andrew Bosworth, the company’s chief technology officer, clarified the reasons behind two specific failed demos in an Instagram AMA. The issues arose during an influencer’s request for cooking guidance from the AI assistant and Mark Zuckerberg’s attempt to answer a WhatsApp call.
Bosworth explained that when the chef prompted the device by saying, “hey Meta, start Live AI,” it triggered every single Meta Ray-Ban’s Live AI within the premises. “We had routed Live AI traffic to our development server to isolate it, but that routing unintentionally affected everyone on those access points in the building. We effectively launched a DDoS attack on ourselves,” he stated, countering Zuckerberg’s claim that the Wi-Fi was to blame.
The problem with the video call was described by Bosworth as “more obscure.” It involved an unprecedented bug that surfaced when the display glasses went to sleep just as the call notification was received. This particular bug has been addressed, although Bosworth acknowledged that a live demo was not the ideal scenario for such a malfunction to become apparent.
Despite these technical setbacks, Meta’s effort to present an honest live product demonstration stands out amidst the prevalent trend among companies like Apple and Google of using pre-recorded videos and generative AI showcasing errors.
0 Comments
