Written by 10:09 pm AI, Discussions, Uncategorized

**Google VP Confirms Authenticity of Gemini AI Demo Video**

Speaking at Fortune’s Brainstorm AI conference in San Francisco on Monday, Sissie Hsiao, Goog…

Key stakeholders are engaged in intense competition for dominance in the dynamic AI market. Google recently introduced teaser videos showcasing their latest Gemini big vocabulary model. Unlike their usual cautious approach, Google took a more proactive stance in this unveiling. However, the revelation that the video was not a real-time demonstration of the AI in operation cast a shadow over the event.

The featured AI model in the Google videos exhibits impressive bidirectional capabilities, adept at interpreting and managing information from both video and audio sources. This development marks a significant milestone for Google, particularly amidst stiff competition from rivals like OpenAI. Reports from Bloomberg indicated that the video was generated using still frames from a film, rather than real-time processing of audio and video inputs.

During Fortune’s Brainstorm AI conference in San Francisco, Bard Sissie Hsiao, Google’s Vice President and General Director of Google Assistant, addressed the controversial video presentation. She highlighted the benchmarks set by Gemini and its potential impact on Google’s future endeavors.

Hsiao affirmed the authenticity of the video, stating that the design actions and prompts were genuine, albeit some parts were condensed for conciseness. The demo showcased the AI model’s multimodal capabilities, including the recognition of various elements like blue lines, curved lines, and a duck drawing, providing real-time information and solutions related to the drawings.

The demonstration underscored Gemini’s proficiency in handling diverse challenges, ranging from high school physics problems to complex ethical dilemmas. Despite Gemini Ultra’s scheduled release next year, early reports suggest its superiority over OpenAI’s GPT-4 in the majority of performance benchmarks. Currently, Bard utilizes the less advanced Gemini Pro, akin to GPT 3.5.

Hsiao emphasized that these Gemini models will play a pivotal role in enhancing Google search functionality and the popular Google Bard chatbot, lauded as “the most preferred free chatbot currently available.”

Visited 17,217 times, 1 visit(s) today
Last modified: February 7, 2024
Close Search Window
Close