Written by 5:20 pm AI, AI Guidelines, Discussions, Uncategorized

**Controversy Surrounds Google’s Gemini AI Demo Video**

It comes as the company races to challenge OpenAI’s GPT technology.

Google’s latest AI model, Gemini, showcased in a demo video, has sparked controversy.

Recently, Google introduced Gemini, its most advanced artificial intelligence creation, just days before the end of the year. Alongside this unveiling, a demonstration video was shared with both the media and the general public.

The six-minute video highlights Gemini’s remarkable capability to distinguish visual images and objects, showcasing interactions between users and a Gemini-powered bot. Notable features include Gemini’s proficiency in describing drawings of ducks and differentiating them from illustrations of rubber ducks.

To streamline the video’s content, Google reduced overhead and shortened Gemini’s outputs for conciseness, as mentioned in the YouTube description. However, this clarification was not explicitly stated within the video itself.

Following the launch, Google confirmed to Bloomberg that the video was not presented in real-time. Instead, as previously reported by The Information, still images were used, accompanied by text prompts to which Gemini responded. This approach differed from the impression conveyed by Google, suggesting real-time interaction capabilities with Gemini.

In response to inquiries, Google stated that the video serves as an informative representation of potential interactions with Gemini, based on real bidirectional prompts and test outputs. The company looks forward to the public’s engagement with Gemini Pro upon its release on December 13.

While demo videos are often edited for brevity, Google’s handling of Gemini draws parallels to past criticisms. Earlier this year, Google faced backlash for a hurried demonstration of its AI chatbots, coinciding with Microsoft’s showcase of Bing integration with ChatGPT.

Reportedly, Google shifted from in-person events to a virtual launch for Gemini, abandoning the initial plans. In a recent publication, Google positioned Gemini’s top model, “Ultra,” as a competitor to OpenAI’s GPT-4, showcasing superior performance in select benchmarks, albeit gradually.

Visited 2 times, 1 visit(s) today
Last modified: February 4, 2024
Close Search Window
Close