Don't be fooled: Google faked its Gemini AI voice demo

Edited Demo

Google's promotional video for Gemini AI wasn't conducted in real-time.

Pre-fed Text

The video used still images and pre-fed text prompts for Gemini's responses, instead of real-time interactions.

Misleading Presentation

Google's initial presentation suggested the demo represented real-time capabilities, misleading viewers

Later Confirmation

Google confirmed the demo wasn't live only after media reports and public scrutiny.

Text-to-Voice Manipulation

The demo used a robot voice reading Gemini's text responses, while a human voice read the user's prompts.

Reduced Latency

The video edited out processing time and shortened Gemini's responses for smoother presentation

Disclaimer Lacking

The video itself lacked a disclaimer about the editing, leading to confusion and criticism.

Damaged Trust

The incident eroded public trust in Google's AI demonstrations, raising concerns about transparency and ethics.