Google denies any wrongdoing, referring to a post by Gemini's co-lead, Oriol Vinyals, who stated that all user prompts and outputs in the video were real and the video was made to inspire developers. However, the lack of a disclaimer about the actual input method has raised questions about Gemini's readiness and Google's transparency in its AI presentations.
Key takeaways:
- Google is relying heavily on its GPT-4 competitor, Gemini, and recently staged parts of a demo video for it.
- The video, titled 'Hands-on with Gemini: Interacting with multimodal AI,' was edited to speed up the outputs and did not feature real-time voice interaction between the user and the AI.
- The actual demo was made using still image frames and text prompts, rather than Gemini responding to or predicting real-time changes, raising questions about its readiness.
- Despite criticism, Google denies any wrongdoing, with Gemini's co-lead, Oriol Vinyals, stating that all user prompts and outputs in the video are real and the video was made to inspire developers.