Google experienced a cringe-worthy moment during its much-anticipated Made by Google event when its AI model, Gemini, failed during a live demo.
Story Summary:
- Gemini failed twice during a live demonstration, unable to provide the desired result.
- The AI model was being showcased as Google’s most powerful yet.
- Social media users quickly noticed and reacted to the embarrassing mishap.
Watch: Gemini Live demo suffers epic fail
During the event, Google aimed to showcase the advanced capabilities of its Gemini AI model, particularly its integration within the Calendar app.
However, when prompted to perform a task, Gemini repeatedly failed, returning to the initial prompt instead of providing the requested information.
The demo malfunctioned twice, leading to an awkward and disappointing moment for Google, especially as the event was live-streamed.
Google has been promoting Gemini as its most powerful AI model to date, with features designed to offer a more consistent and emotionally expressive interaction experience.
Despite these claims, the live demonstration did not go as planned, and the glitches did not go unnoticed by viewers.
Gemini Live, which is rolling out on Tuesday following its announcement at Google’s I/O 2024 developer conference, is intended to rival the recently launched Advanced Voice Mode for OpenAI’s ChatGPT.
The feature allows users to have in-depth voice chats with Gemini on their smartphones, with Google emphasising its enhanced speech engine and improved conversational abilities.
A Google spokesperson highlighted the model’s capability, stating:
“Live uses our Gemini Advanced models that we have adapted to be more conversational,” adding that its large context window enables it to handle long conversations effectively.
However, the live demo mishap overshadowed these advancements, drawing attention instead to the model’s shortcomings.
