Google’s Gemini AI Demo Faces Criticism for Allegedly Being “Fake”

By Roze 2 Min Read
Google's Gemini AI Demo Faces Criticism for Allegedly Being Fake

Google’s recent “hands-on” tech demo showcasing its new artificial intelligence model, Gemini, has come under fire from critics who claim that the demonstration was “basically entirely fake.” The video, which gained significant attention with 2.1 million views on YouTube, displayed Gemini seamlessly interacting with a human operator in what appeared to be real-time scenarios. However, critics argue that the interactions were misleadingly edited, creating an unrealistic portrayal of Gemini’s capabilities.

Oriol Vinyals, an executive at Google DeepMind, clarified that while the user prompts and outputs were real, they were “shortened for brevity.” The interactions were, in reality, text-based and took longer than represented in the edited video. Google uploaded the video with a disclaimer acknowledging the reduction of latency and the shortening of Gemini outputs for brevity.

Criticism emerged on social media, with accusations that Google lied and faked the AI demo. Some software developers and Google employees expressed dissatisfaction with the way the video presented Gemini’s capabilities, highlighting the need for more transparency in AI demonstrations. One Google employee suggested that the video portrayed an unrealistic image of Gemini, emphasizing the ease of making the AI tool appear more advanced than it truly is.

Despite the controversy, Google claims that Gemini surpasses leading AI models in various benchmarks related to reasoning, math, language, and other metrics. Launched as a competitor to OpenAI’s ChatGPT, Gemini aims to excel in multiple aspects, including outperforming GPT-4 in specific benchmarks.

You may be interested in our other articles about technology if you enjoyed this one:

Share this Article
Leave a comment