In an opinion piece that was released in Bloomberg, Google said that it had staged parts of its demo video for Gemini, a new competitor to the GPT-4. The company changed both the speed of output and the voice that sounds like a person or an AI user.
Gemini didn’t react in real time to a drawing or change in the items on the table in the first demo. Instead, it was made “using still image frames from the footage and prompting via text.” The false movie makes people doubt Gemini’s abilities, especially since the demo didn’t have a disclaimer.
On X, Oriol Vinyals, who is also the co-lead for Gemini, denied doing anything wrong and said that “all the user prompts and outputs in the video are real.” He also said 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