SAN FRANCISCO — OpenAI released a major update to ChatGPT this week, merging its Advanced Voice Mode with the standard text interface to ensure the AI can misunderstand you in two formats at the exact same time.

The update removes the dedicated voice screen, allowing the AI to generate long text responses while simultaneously speaking over the user. This approach mimics the experience of a frantic Zoom meeting where a colleague keeps typing in the chat while refusing to mute their microphone. “We wanted to capture the authentic panic of modern communication,” said Elena Ross, OpenAI’s Director of Conversational Chaos. “Now, ChatGPT can verbally explain why you’re wrong while visually bullet-pointing your errors, ensuring you never miss a correction.”

Early adopters report the integrated interface creates a unique “sensory wall” where the AI verbally apologizes for a mistake while textually doubling down on the incorrect answer. The system can now flash code snippets on screen while humming, a feature designed to simulate the distraction of working in an open-plan office. “It’s about density,” noted Marcus Thorne, Vice President of Overstimulation Strategy. “Why just hear the lie when you can read it in real-time? We are giving users the complete package of confusion.”

At press time, OpenAI announced a future update that will vibrate the phone and flash the camera light to ensure the user is paying absolute attention to the wrong answer.

Inspired by actual events.