Gemini AI Faces Mega Backlash, Google Makes Troubling Admission

Technology company and Artificial Intelligence (AI) pioneer Google has faced a lot of criticism in the last few days following the release of a demonstrative video that detailed the capabilities of Gemini, its new multimodal AI model.

Google Floats Gemini AI

After teasing the public about the potential launch of the Large Language Model (LLM) at the I/O developer conference that was held in June, the tech giant finally unveiled the Gemini AI earlier this week. Sundar Pichai, Google CEO shared a video on the X app touting the potential of the AI model, with emphasis on its interactive capability.

In this video which looked like an interaction between a human and a voice assistant AI, Gemini accurately identified a duck that was created using lines and shapes. The video suggested that Gemini AI can recognize visual pictures and physical objects and even go as far as telling their difference. 

This unique feature demonstrated a new frontier in AI design and modeling. Gemini AI underscored a more integrated use of LLM that is likely to elevate excellent user experience, making it a tough competition for OpenAI’s ChatGPT. In addition, it seemed like Google was trying to inform the public that with Gemini, they could have a smooth voice conversation with an AI.

Unfortunately, the six-minute video sparked some negative reactions from members of the public who felt that the content was doctored. It is worth noting that the company added a description on YouTube stating that “For the purposes of this demo, latency has been reduced, and Gemini outputs have been shortened for brevity.” 

However, the disclaimer was not mentioned when the demonstrative video was shared on the X app.

Google Admits Demo Video Was Edited

Google has openly admitted to Bloomberg that the demo was not conducted in real time. 

However, the company said that “the video is an illustrative depiction of the possibilities of interacting with Gemini, based on real multimodal prompts and outputs from testing.”

To achieve the six-minute demo, the Google team behind the development of the AI, used still images and fed text prompts that Gemini responded to. This is unlike the notion which Google led viewers to believe.

Despite the backlash received, Google said it is still looking “forward to seeing what people create when access to Gemini Pro opens on December 13.”