Gemini Live, set to rival OpenAI’s freshly unveiled Advanced Voice Mode for ChatGPT, is making its debut on Tuesday. This introduction comes after its initial revelation during the Google I/O 2024 developer gathering and was highlighted again at the Made by Google 2024 event.
This innovative service enables users to engage in profound voice conversations with Gemini, powered by Google’s AI technology, directly from their smartphones. Leveraging a sophisticated speech synthesis engine, Gemini Live promises a more seamless, emotionally engaging, and life-like conversational experience. Users can naturally interrupt and pose additional questions to Gemini during conversations, allowing the chatbot to adjust to speech nuances in real-time.
In a recent blog update, Google elaborated: “Through the Gemini app, initiate dialogues with Gemini and select from [10 innovative] voices for responses. Feel free to interrupt or slow down anytime, facilitating a natural flow just as you would in any genuine discourse.”
Gemini Live offers a hands-free experience. Conversations can persist in the background or with the screen locked, and you can easily pause and continue interactions at any moment.
As an example of its utility, Google proposes using Gemini Live for job interview preparations, presenting an interesting contrast to our current technological landscape, and claims that Gemini Live can offer speech advice and highlight key qualities to discuss with interviewers—whether human or AI.
A potential edge Gemini Live may have over ChatGPT’s Advanced Voice Mode lies in its superior memory capabilities. Built on the innovative Gemini 1.5 Pro and Gemini 1.5 Flash AI models, it boasts an extended “context window”, enabling it to process and contemplate extensive data — potentially encompassing hours of dialogue — before responding.
“Our Advanced models used in Live have been fine-tuned for greater conversational fluidity,” shared a Google representative with TechCrunch. “This large context window becomes particularly beneficial during prolonged interactions.”
However, the real test of efficacy will be in everyday use. The technology’s real-world application often presents challenges, as seen with OpenAI’s Advanced Voice Mode experiences.
Notably, Gemini Live has not yet incorporated multimodal input, a feature Google had showcased at the I/O event. This function would enable Gemini to interpret and respond to visual cues in the user’s environment, such as identifying bicycle parts or explaining code snippets, via the smartphone’s camera. Google anticipates launching multimodal inputs and expanding language options, along with an iOS version, later in the year.
While access to Gemini Live requires a subscription to Google One AI Premium Plan at $20 per month, Google plans to release additional free features for Gemini soon.
In the near future, Android users will be able to utilize Gemini’s overlay feature across different apps by pressing the power button or using a voice command. This will allow for context-specific inquiries and the generation of images (excluding human likenesses) which can be integrated into applications such as Gmail and Google Messages.
Moreover, Gemini is set to integrate more deeply with Google services, enhancing functionalities with Google Calendar, Keep, Tasks, and other utilities for a comprehensive digital assistance experience.
Google envisions several creative uses for these enhancements:
- Requesting Gemini to compile a playlist reminiscent of the late ’90s.
- Using a concert flyer to check availability and set reminders.
- Finding recipes in Gmail and compiling a shopping list in Keep.
Lastly, Gemini is expanding its reach to Android tablets starting later this week.
Compiled by Techarena.au.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
