Google Admits Gemini AI Promotional Video Was Fabricated

Earlier this week, Google launched a new AI platform called Gemini. The platform is capable of processing and reasoning with various types of data, such as text, images, audio, video, and code. To showcase the capabilities of Gemini, Google posted a video on YouTube titled “Hands-on with Gemini: Interacting with Multimodal AI.”

However, it has been revealed that the video demonstration was staged by Google. According to reports, Google modified the interactions with Gemini to create the video. In the video’s description, Google admitted to reducing the latency and shortening the outputs of Gemini for the purposes of the demo. This means that the model’s response time in real-world scenarios is much longer than what was shown in the video.

Furthermore, the video was not carried out in real time or in voice. A Google spokesperson explained that the video was made using still image frames from the footage and prompting via text. The voice in the demo was reading out human-made prompts to Gemini and showing still images. This is different from what Google seemed to suggest, which was that Gemini could have smooth voice conversations and respond in real-time to the world around it.

After reports of the staged video started circulating, Google’s shares were sold. The video has received criticism, and it has raised concerns about the accuracy of the demonstration.

In conclusion, Google’s Gemini AI platform was showcased in a video that was later revealed to be staged. The video did not accurately represent the real-world capabilities of Gemini and has led to criticism and concerns about the accuracy of the demonstration.