In an opinion piece published in Bloomberg, Google has admitted to staging parts of its demo video for newly launched GPT-4 competitor, Gemini. The company not only edited the speed of output, but also added the implied voice between human voice and AI user.
The original demo was made “using still image frames from the footage, and prompting via text,” rather than having Gemini respond to a drawing or change of objects on the table in real time. The misleading video puts into question Gemini’s capabilities, especially since the demo lacked a disclaimer.
Gemini’s co-lead, Oriol Vinyals, on X denied any wrongdoing and said that “all the user prompts and outputs in the video are real,” and that his team made the video “to inspire developers.”
Really happy to see the interest around our “Hands-on with Gemini” video. In our developer blog yesterday, we broke down how Gemini was used to create it. https://t.co/50gjMkaVc0
We gave Gemini sequences of different modalities — image and text in this case — and had it respond… pic.twitter.com/Beba5M5dHP
— Oriol Vinyals (@OriolVinyalsML) December 7, 2023
